Effects of Sport Related Concussion on Academic ... · Effects of Sport-Related Concussion on...
Transcript of Effects of Sport Related Concussion on Academic ... · Effects of Sport-Related Concussion on...
Effects of Sport Related Concussion on
Academic Performance in High School Athletes
A Thesis
Submitted to the Faculty
of
Drexel University
by
Ginger Ann Stringer
in partial fulfillment of the
requirements for the degree
of Doctor of Philosophy
October 2014
© Copyright 2014 Ginger Ann Stringer. All Rights Reserved
ii
ACKNOWLEDGEMENTS
I would like to acknowledge the support I have received from numerous
people in completing this dissertation. First, many, many thanks to my advisory
committee for this dissertation. My advisor, Dr. Yvonne Michael has offered
continued support and always-thoughtful feedback and encouragement. Without
her guidance, this dissertation would not have come to fruition.
A huge thanks to Dr. Craig Newschaffer for his support, feedback, and for
encouraging my research in a topic outside the bounds of the department faculty.
Without his introduction to Dr. Andrew Lincoln, I would not have had access to
the data or support necessary for this research. Dr. Lincoln has been
instrumental and supportive in lending his expertise in sport-injury epidemiology
research. I would also like to thank Dr. Jennifer Taylor and Dr. Ed Gracely for
their insight and feedback at various stages of my research. Special thanks to Dr.
Terry Hyslop, a great friend and mentor, who offered much professional and
personal support and advice during my graduate studies.
Finally, I would like to thank my family for their support as I pursued my
education – my mom, Jill, Athena, Liam, Fern, and Lisa. My mother, Paula
Shimmin, has always encouraged my curiosity and been a great supporter of my
academic goals. Most of all, thank you to my wife, Lisa Rosen, for patiently
supporting my research and educational goals and letting me make them a
priority, sometimes at the expense of things I know we would have both rather
been doing.
I could not have done this without any of you.
iii
TABLE OF CONTENTS List of Tables ........................................................................................................... v List of Figures ........................................................................................................ vi Abstract ................................................................................................................. vii 1. DISSERTATION OVERVIEW ............................................................................. 1 1.0 Concussion Research Background and Definition ..................................... 2 1.1 Concussion in Adolescent Athletes .............................................................. 3 1.2 Concussion Assessment and Diagnosis ...................................................... 11 1.3 Academic Impact of Sport-Related Concussion ......................................... 12 1.4 Dissertation Aims ....................................................................................... 14 1.5 Dissertation Organization ........................................................................... 15 2. LITERATURE REVIEW: POST-ACUTE COGNITIVE EFFECTS OF SPORT RELATED CONCUSSION IN HIGH SCHOOL AND COLLEGIATE ATHLETES ................................................................................... 17 Abstract ............................................................................................................. 18 2.1 Introduction ............................................................................................... 20 2.2 Methods ..................................................................................................... 23 2.3 Results ....................................................................................................... 26 2.4 Discussion .................................................................................................. 34 2.5 Tables and Figures .................................................................................... 38 3. RELATIONSHIP BETWEEN CONCUSSION INJURY OR LOWER LEG INJURY AND ACADEMIC PERFORMANCE IN HIGH SCHOOL ATHLETES ............................................................................... 46 Abstract ............................................................................................................. 47 3.1 Introduction ............................................................................................... 49
iv
3.2 Methods ...................................................................................................... 51 3.3 Results ....................................................................................................... 56 3.4 Discussion .................................................................................................. 59 3.5 Tables and Figures ..................................................................................... 65 4. LONGITUDINAL ANALYSIS OF ACADEMIC PERFORMANCE IN HIGH SCHOOL STUDENTS WITH ATHLETIC INJURIES ......................................... 70 4.0 Abstract ...................................................................................................... 71 4.1 Introduction ................................................................................................ 73 4.2 Methods ...................................................................................................... 74 4.3 Results ....................................................................................................... 80 4.4 Discussion ................................................................................................. 83 4.5 Tables and Figures .................................................................................... 86 5. CONCLUSIONS AND RECOMMENDATIONS .............................................. 90 5.1 Summary of Findings .................................................................................. 91 5.2 Public Health Implications ....................................................................... 94 5.3 Policy Implications .................................................................................... 95 5.4 Limitations of Current Research ................................................................ 97 5.5 Future Research Considerations ............................................................... 98 LIST OF REFERENCES ..................................................................................... 100 VITA .................................................................................................................... 113
v
LIST OF TABLES
2.1 Summary of included study demographics, methodology, and primary findings ................................................................................................................. 39 2.2 Summary of studies included in meta-analyses reporting neurocognitive results from IMPACT ............................................................................................ 40 2.3 Standardized mean differences for ImPACT cognitive domains .................... 41 3.1 Distribution of injury by sport ........................................................................ 65 3.2 Severity of concussion injury by sport ........................................................... 66 3.3 Distribution of injury by grade level ............................................................... 67 3.4 Change in school attendance by injury type .................................................. 68 3.5 Distribution of concussion injury by gender and sport ................................. 69 3.6 Odds of injury severity by gender ................................................................... 69 4.1 Distribution of injury in the study population ............................................... 86 4.2 Comparison of mean GPA in the academic period prior to injury to mean GPA in the academic period of injury ........................................................................... 87 4.3 Effect of injury type and severity on mean GPA ............................................ 87 4.4 Results of random intercept models .............................................................. 88 4.5 Results of random intercept model restricted to lower grades ...................... 88
vi
LIST OF FIGURES
2.1 Schematic illustrating the process of selecting publications for systematic review .................................................................................................................... 38 2.2 ImPACT Total Symptom Score forest plot ..................................................... 42 2.3 ImPACT Verbal Memory Score forest plot ..................................................... 43 2.4 ImPACT Visual Memory Score forest plot ..................................................... 43 2.5 ImPACT Processing Speed Score forest plot .................................................. 44 2.6 ImPACT Reaction Time Score forest plot ...................................................... 45 3.1 Severity of concussion injury by sport ............................................................ 67 3.2 Severity of concussion injury by sport and gender ........................................ 68 4.1 Growth curve model illustrating change in mean GPA over time ................. 89
vii
ABSTRACT Effects of Sport-Related Concussion on Academic Performance
in High School Athletes Ginger A. Stringer
Yvonne Michael, ScD Background. Concussive impacts to the head are a natural part of most athletic
competition. Even in sports that do not permit player contact, incidental contact
that can result in concussion often occurs. This dissertation addresses the general
hypothesis that sport related concussions have a negative impact on academic
performance in high school athletes.
Study Design. A systematic review of current literature on the post-acute
neurocognitive effects of concussion provides an introduction to the importance
of this study. A retrospective cohort of athletes participating in school sports in a
public school district is assessed for changes in school attendance and academic
performance following concussion compared to ankle or leg injury.
Methods. The systematic review included meta-analytic methods. Analysis of
the retrospective cohort included longitudinal graphical and multivariable
random coefficient methods that characterized changes in academic performance
over time in high school athletes with and without concussion. Random
coefficient models were used to assess effect modification by predictor variables
including gender and age.
Results. Researchers have shown that adolescent athletes experience a range of
neurocognitive symptoms six or more days following concussion injury. In the
retrospective cohort analyses, athletes with concussion injuries have greater odd
viii
of being grade 9 (OR=2.2, p<0.001, 95%CI 1.5-3.2) or grade 10 (OR 2.1, p<0.001,
95%CI 1.5-3.1) than grade 12 and had greater odds of an increase in the number
of days absent from school (OR=1.7, p<0.011, 95%CI 1.3-2.2) when compared to
athletes with ankle or leg injury. Analyses of a subgroup of younger athletes
(grades 9 and 10) with more severe injuries indicate a 16% greater decline in
academic performance for the group with concussion injury than those with
ankle or leg injury (coeff= -0.18, p=0.017, 95% CI= -0.32 to -0.03).
Conclusions. These analyses suggest that young athletes with concussion
injury are at risk for decline in academic performance. Increased absence from
school results in decreased instruction time, social isolation, and other factors
that may negatively impact academic achievement. Future research is warranted,
examining academic performance in high school athletes with concussion injury.
1
CHAPTER 1: DISSERTATION OVERVIEW
2
1.0 Concussion Research Background and Definition
Concussive head impacts are a natural part of most athletic competition.
Even in sports that do not permit player contact, incidental contact with other
players, equipment, or the playing surface that may result in injury. Sport-related
concussions, a subgroup of mild traumatic brain injury (MTBI), are accompanied
by a constellation of poorly defined short and long-term physical and cognitive
consequences for the athlete. In 2001(1), 2004(2), 2008(3), and 2012(4) several
sport governing bodies convened international symposia to develop a consensus
statement and recommendations regarding the safety and health of athletes who
suffer concussion injuries. The agreement statements from this group provide a
concise and consistent definition of concussion, a description of clinical issues,
assessment tools, and return to play protocol. The consensus definition of sport-
related concussion is “a complex pathophysiological process affecting the brain,
induced by a traumatic biomechanical force.” (1-4) This definition is
accompanied by several common symptoms that are used in defining the severity
of a concussion.
Concussions are characterized by acute and long-term changes in the brain
resulting in a variety of physical (e.g. balance problems, loss of consciousness),
cognitive (e.g. slowed reaction time, memory loss, concentration difficulty), and
emotional (e.g. emotional lability) symptoms.(3, 5) These symptoms may present
immediately at the time of injury, or appear in the hours and days following the
injury.(6-10) The clinical symptoms used to diagnose concussion have been
described in detail(2, 3, 6, 9, 11-16). The acute symptoms that have been
3
validated as prognostic of poor injury progression and slower recovery include
amnesia, loss of consciousness, headache, dizziness, blurred vision, attention
deficit, and nausea.(14, 17, 18) Additional subjective symptoms have been
described, but have not been proven as prognostic features of concussion
recovery.(18)
1.1 Concussion in Adolescent Athletes
Participation in structured activity during adolescence is associated with
positive psychological, social, and educational outcomes.(19-21) The structure
provided by athletics and other extracurricular activities allows for an
appropriate expression of normal adolescent urges such as risk-taking and
sensation-seeking behavior that is otherwise uncontrolled and potentially
dangerous.(22, 23) Additionally, participation in structured activity encourages
development of positive time management and leadership skills(23-28) and
engaging in sports and other athletic endeavors expose adolescents to
experiences that increase confidence and competence in problem solving.(29)
Participation in high-school athletics is predictive of lower substance abuse rates,
increased probability of graduation from college, and higher occupational
achievement than non-participants.(19, 30)
Despite the positive health and social outcomes associated with physical
activity and sports participation, there is an assumed risk for the athletes.
Recently, intense, year-round, single-sport training has become increasingly
common in younger athletes who have aspirations of competition at the highest
4
level.(30, 31) This training often does not accommodate the quickly changing
physical structures of adolescence, and leads to increases in injury related to
overuse and overtraining.(30, 31) Sport-related injury negates some of the
positive emotional and social benefits of participation. Athletes with injury are
more likely to show depressive symptoms and experience social isolation than
non-injured athletes.(32)
Adolescents who chose to participate in athletics and other structured
programs are different in many ways than those who do not. They are more
likely to be of higher social economic class(20, 32, 33), have a more positive
orientation toward education (resulting in better academic performance)(20, 22,
32, 34), and be of nonminority ethnicities(20, 22, 34) than those who do not
participate in structure athletic programs. These characteristics make research of
outcomes related to sports participation difficult, as it is challenging to detangle
the effects of participation from the factors that predispose athletes to
participate.(26) Moderators of the effect of athletic participation on academic
performance include gender, ethnicity, age, and academic environment.(26)
Athletic participation is variable across these subgroups and effects of
participation may also impact these groups differentially.
Sport-related concussion injuries are of particular concern in young
athletes. Approximately 30 million children and adolescents in the United States
participate in sports every year.(33, 35) Surveillance in emergency departments
between 2001 and 2005 indicates that among patients examined for sport-related
concussion, a majority were pediatric patients. The highest rates were seen in
5
young athletes aged 10 to 19.(33) Given the physical and cognitive changes
occurring during puberty, adolescents are particularly vulnerable for concussion
injuries. Injury surveillance and other research consistently find the highest
rates of adolescent sport-related concussion in football, wrestling, soccer, and
basketball.(36-40)
1.1.1 Incidence of Concussion in Adolescent Athletes
In the United States, over 7.5 million adolescents participate in high
school athletic programs each year.(41) American football, alone, accounts for 1.1
million high school participants every year.(41) The incidence of sport-related
concussion varies greatly by sport and performance level, and has been
historically hard to estimate. The Centers for Disease Control incidence estimate
of approximately 300,000 head injuries attributable to high school athletics
every year is widely referenced in the literature. (42) Recent studies of large,
nationally representative samples that include athletes in a wide range of sports
are beginning to offer a better estimate of the true injury burden in adolescent
athletes. Extrapolation from US Census data and estimates of incidence from
several studies to account for underreporting and undiagnosed injury produce
rates of up to 300 concussion injury per 100,000 for children age 15-19
participating in athletics.(43)
Two recent longitudinal studies have produced an estimated concussion
rate of approximately 2.5 injuries per 10,000 athletic exposures (AE)(44, 45),
with concussions representing up to 15% of reported athletic injuries. Athletic
6
exposure (AE) are defined as one athlete participating in one practice or
competition. Both studies, one a longitudinal study of athletes in a large public
high school system over ten years(44) and the other a study of a national sample
of high school athletic injury data during the 2008-2010 academic years(45),
indicate that while overall rates of concussion are higher in boys than in girls,
with a majority of concussions resulting from football, risk of concussion is
higher in girls than in boys in gender-comparable sports.(44, 45) Overall, the rate
of concussion injuries in high school athletics is increasing.(44, 46) This trend
could be due to increased awareness of the consequences of concussion injuries,
improvements in diagnostic processes, and/or an actual increase in the number
of injuries. Since concussion analysis, management, and diagnosis became a
priority in the 1990’s, there has been an overall decrease in serious brain injuries
in most sports.(44, 47)
1.1.2 Predictors of Concussion Severity and Prolonged Recovery
As early as the 1960’s, researchers noted that the effects of even the
mildest concussion may linger for an extended period, or even be
irreversible.(48, 49) There are competing schools of thought regarding persistent
symptoms following concussion. Whether the symptoms are a direct
consequence of the initial injury or manifestations of psychological or emotional
after effects, there is increasing concern that these symptoms represent a largely
unnoticed and untreated syndrome resulting in significant disability.(18) There
is insufficient data available to quantify the proportion of injuries or individual
7
characteristics that result in lingering post-concussive symptoms. However,
female gender, younger age, history of concussion, and presentation of acute
concussion symptoms appear to be predictive of injury severity and prolonged
recovery from the injury.
Adolescent brains do not follow the same general recovery pattern as
adults following concussion injuries. While adults generally resolve symptoms
within the first week after injury, recent research indicates that young athletes
may begin showing new cognitive symptoms 7-10 days following injury.(33, 35)
The adolescent brain may be more sensitive to metabolic changes following
injury than adult brains. This imbalance may impact recovery time, partially
explaining the delayed recovery trajectory observed in adolescents compared to
adults.(33, 50) History of previous concussion also increases recovery time in
adolescents. Subsequent concussions demonstrate more severe immediate
symptom presentation and are associated with long-term neurological
deficits.(33, 36, 51)
1.1.2.1 Gender
Female athletes appear to be at greater risk for concussion and experience
a longer duration of symptoms than male athletes in gender-comparable
sports.(52, 53) A recent prospective cohort study describes sex differences in
memory, postural stability, and symptom scores following concussion, with
female athletes performing significantly worse than male athletes up to a week
following injury.(54) It is unclear, however, if these results are due to an actual
8
disparity in risk and recovery, or if the findings can be explained by other non-
biological factors. For example, female athletes may be more likely to report
symptoms and honestly engage in recovery efforts instead of masking symptoms
to hasten return to play.(55, 56)
1.1.2.2 Age
While adolescent changes that influence risk of athletic injury are most
visible in physical transitions in body shape and size, concurrent change in
cognition, emotion, and social development also influence risk of injury and can
be adversely impacted as a result of injury.(27, 57) Change in brain structure and
function during this phase is not readily apparent, but occurs at a rapid rate.
During puberty, brain maturation is characterized by reduction of neuronal axons
and increased myelination.(22) It is estimated that during adolescence synapses
can be generated and lost on the magnitude of 30,000 per second.(23)
The changing adolescent brain is particularly vulnerable to injury. During
adolescence, shifting activation of frontal regions of the brain are associated with
changes in cognitive control.(57) Risk taking behavior and emotional reactivity
are associated with developmental change in the subcortical limbic regions.(22,
57) These changes are nonlinear and highly individual. Damage to changing
structures during development could have lasting consequences on behavior and
cognition. There is evidence that these pubertal changes in brains and hormones
interact with contextual factors (such as injury) to influence academic
achievement.(20, 27)
9
While most cognitive detriments associated with concussion are the same
in children as in adults, the fact that the adolescent brain is actively maturing
may result in significant secondary consequences in children as their ability
perform day-to-day tasks impacts their educational and social attainment.(58)
Even in the absence of immediate presentation of clinical symptoms commonly
observed in adult concussion, injury to the adolescent brain occurs at a lower
energy threshold than in adults that manifest in the presentation in symptoms in
the post acute period.(59, 60) It is also suspected that despite normal
neuropsychological presentation following sport concussion, some children may
suffer a variety of behavioral and cognitive sequelae.(59, 61)
The effectiveness of adolescent concussion assessment and diagnosis is
influenced by maturing cognition in children. In early adolescence, healthy
children are rapidly developing choice reaction time, working memory, and new
learning(58). This has implications for baseline and post-concussive cognitive
assessments as the developmental changes are of a comparable magnitude to
post-concussive impairments, potentially confusing assessment of symptoms(58,
62, 63).
1.1.2.3 History of Concussion
A small study of high school and collegiate athletes by Iverson and
colleagues assessed neuropsychological measures in athletes with three or more
concussions and found significant memory deficits compared to athletes with no
previous concussion.(64) This study was small (n=38), but provides convincing
10
preliminary evidence of significant cumulative effect of multiple concussion in
adolescent athletes using a validated, computerized assessment tool. The
researchers chose a sample of athletes with multiple concussions in an effort to
replicate and expand upon previous research that found athletes with 3 or more
concussions present increased concussion-like symptoms at baseline than
athletes without concussion.(65-67)
In an attempt to characterize the prolonged effects of concussion in high
school athletes, a study in New Jersey found that athletes who had sustained
recent concussions had poorer concentration and memory function than athletes
with no history of concussion.(68) The study collected neurophysical
information on 223 athletes in three groups: those with no concussion history,
symptom free athletes with one or more concussions more than six months prior
to testing, and those who experienced concussion 1 week prior to testing.
Athletes with history of two or more concussions and those with recent
concussion demonstrated poorer attention and concentration than athletes with
no concussion history. Additionally, the researchers found that athletes with
recent concussions and currently asymptomatic athletes who had a history of two
or more concussions had lower cumulative grade point averages than athletes
without concussion.(68) The correlation between low GPA and concussion
incidence must be studied in a larger sample with longitudinal data collection
before assertions can be made.
Existing literature documents cognitive decline following concussion, but
is limited by application of neurocognitive tests that are not predictive of long-
11
term function. Additionally, most studies have had small sample size and do not
follow athletes over long periods to assess function in the post-acute period
following injury.
1.1.2.4 Other prognostic factors
Several characteristics of acute injury have been validated as prognostic of
injury progression and recovery include amnesia, loss of consciousness,
headache, dizziness, blurred vision, attention deficit and nausea.(14, 17, 18)
Additional subjective symptoms have been described, but have not been proven
as prognostic features of concussion recovery.(18)
1.2 Concussion Assessments and Diagnosis
Diagnosis of concussion in sports is complicated on many levels.
Assessment of injured athletes is hindered by the inherent challenges of
diagnosing a condition that relies largely on self-report of symptoms.(11, 69) It
is not uncommon for an athlete to hide or lie about symptoms to avoid being
“benched” for a seemingly minor injury, opting instead to “play through the
pain.”(11, 69) Additional challenges to diagnosis and assessment of symptoms
include stress(11, 33, 37, 70, 71), learning disabilities(11), and hormonal
fluctuations(11, 72) that affect response to concussion assessment queries.
No universally accepted concussion symptom assessment tool has been
developed for diagnosing concussion at the time of injury. Many have been
proposed and are used in practice.(2, 3, 6, 11-16, 73) In 2004, the International
12
Conference on Concussion in Sport published the Sideline Concussion
Assessment Tool (SCAT), which offers patient education in addition to a
mechanism for assessing cognitive features, typical symptoms, and physical signs
of concussions.(2) SCAT was developed through a collaborative and iterative
process consolidating and standardizing existing tools.(2) The SCAT tool has
been revised several times, including a version to accommodate assessment of
concussion symptoms in children.(4) Another frequently used tool is the
Standardized Assessment of Concussion (SAC) assessment tool(16) for on-field
assessment. The SAC is a series of questions in four sections that evaluate areas
of orientation, memory, concentration and delayed recall. The test is
administered immediately following injury and the cumulative score is compared
to a baseline value. Multiple versions of the evaluation are available to minimize
the potential for practice effects. Deficits of 1 or more points on a 30 point scale
are indicative of impaired cognitive function.(6, 16) Additional assessment tools
are in use, and all rely on measures of memory, balance, and self-report of
symptoms. All of these tools are useful in initial assessment of injuries and can
inform clinical management during recovery. Additional research regarding the
predictive value of these tools in an athletes recovery trajectory is necessary.(74)
13
1.3 Academic Impact of Sport-Related Concussion
Concussion injuries may have significant impact on academic performance
in high school athletes. In adolescent athletes with concussion, attention must be
paid to balancing the required cognitive rest with academic demands. Minor
attention deficit or impaired information processing can have a significant impact
on an adolescent’s ability to function academically(58) and evidence suggests that
early return to school (before resolution of symptoms) may delay recovery.(33,
75) Studies have noted a lack of research documenting specific outcomes when
assessing the impact of head injuries on scholastic abilities.(76-78) The few
existing studies of academic outcomes focus on severe head injuries and do not
translate to the large number of athletes suffering from concussive symptoms.(77,
79)
For athletes with concussion, cognitive rest may be as important as
physical rest for recovery.(80, 81) Poor communication between healthcare
providers and school administration make it difficult for athletes to get the rest
they need to heal.(9, 33, 71, 81) Assessing the impact of concussions on academic
outcomes will help with future identification of athletes at high-risk for prolonged
recovery and allow educators opportunity for targeted intervention to prevent
negative academic consequences. Currently, many athletes are expected to
return to school environments that require focused attention over long periods
and exposure to loud environments within days of injury.(57, 82, 83)
14
Administrative policies allowing for cognitive recovery periods and
individualized academic rehabilitation plans may allow full recovery from injury,
minimizing academic consequences.
1.4 Dissertation Aims
The aims of this dissertation focus on evaluating the effects on academic
performance of sport-related concussion in high school athletes. The dissertation
provides a systematic literature review of the post-acute cognitive effects of
concussion in adolescents as well as analyses of academic performance in a
population of high school athletes with concussion injury and lower leg injury. A
discussion of the policy implications of the research is included.
Aim 1: Provide a systematic review of the literature regarding the post-
acute neurocognitive effects of concussion in adolescent athletes.
Aim 1a: Use meta-analytical statistical methods to estimate the
overall effects of concussion in observational studies of the post-
acute neurocognitive symptoms in adolescent athletes.
Aim 2: Describe the injury rates and distribution of academic variables in
a population of high school athletes with concussion injury and lower leg
injury.
Aim 3: Characterize the impact of sport-related concussion on academic
performance in high school athletes by describing the functional form of
recorded grade trends over time.
15
Aim 3a: Examine the functional form of academic performance
trends over time in a group of athletes with concussion and a group
of athletes with ankle and knee injuries.
Aim 3b: Estimate the effect of sport-related concussion on academic
performance in high school athletes using random coefficient
models.
1.5. Dissertation Organization
This dissertation contains five chapters. This chapter (Chapter 1) provides
an overview of background information relevant to sport-related concussions and
current literature. In Chapter 2, a systematic literature review and meta-analysis
of the post-acute effects of concussion in adolescent athletes is presented.
Chapter 3 describes the injury and academic data from a population of high
school athletes used to achieve the aims of this dissertation. Chapter 4
investigates the longitudinal effects of sport related concussion injury on
academic performance. Finally, the fifth chapter is a summary of the dissertation
and discussion of the findings with emphasis on the policy implications and
directions for further research.
The research presented in this dissertation was designed to increase the
body of information regarding the impact of concussion in adolescent athletes.
Prolonged cognitive impairment following concussion is of concern in adolescent
athletes, but poorly understood in the context of variables that may impact long-
term outcomes. Academic performance in high school influences later academic
16
opportunities and is association with health standing later in life. The analyses
rely on a restricted data set from a large study population for retrospective
analyses of effects. Results from this dissertation provide a foundation for future
research in this area, including prospective analyses and investigation of more
complete population data.
17
CHAPTER 2: LITERATURE REVIEW - POST-ACUTE COGNITIVE EFFECTS OF SPORT-RELATED CONCUSSION IN HIGH SCHOOL
AND COLLEGIATE ATHLETES
18
Abstract
Background. Adolescent athletes do not follow the same pattern of
recovery following concussion injury that adults demonstrate. Adolescents may
have prolonged or new symptoms that appear beyond the post-acute period
following injury. Understanding the recovery of adolescent athletes following
concussion is essential to minimize the consequences for the future health of the
athletes. The aim of this study was to provide a systematic review of the
literature regarding post-acute (six or more days following injury) neurocognitive
effects of concussion in adolescent athletes.
Study Design. A systematic review of current literature addressing
cognitive effects of concussion in adolescents was conducted. Searches were
conducted in PubMed and Ovid to identify studies published in English between
January 1995 and May 2013. Studies were included in the review if
neurocognitive results for high school age subjects were reported six or more
days following concussion injury.
Methods. Results from neurocognitive tests were pooled where possible
to present overall effects. Pooled standardized mean differences were calculated
using Cohen’s method and DerSimonian and Laird random effect weighting with
heterogeneity estimates taken from the inverse-variance fixed effects model.
Where pooling of study results was not possible, results were presented as a
systematic review with no quantitative summary estimate.
19
Results. Twelve studies were included in the systematic review. Eight
were cohort studies reporting results from ImPACT testing (a computerized
neurocognitive test of concussion symptoms), four studies reported results of
other neurocognitive tests. Adolescent athletes experience a range of
neurocognitive symptoms six or more days following concussion injury. Athletes
in the reviewed studies demonstrated declines in the domains of reaction speed,
memory, and total symptom scores at the time of follow up.
Conclusions. Many domains of cognition remain impaired in adolescent
athletes in the post acute period (six or more days after injury). These findings
support the need for further study of cognitive effects of concussion in
adolescents following the acute period of injury rehabilitation.
20
2.1 Introduction
Concussive head impacts are a natural part of most athletic competition.
Even in sports that do not permit player contact, incidental contact with other
players, equipment, or the playing surface that may result in injury. Sport-related
concussions, a subgroup of mild traumatic brain injury (MTBI), are accompanied
by a constellation of poorly defined short and long-term physical and cognitive
consequences for the athlete. In 2001(1), 2004(2), 2008(3), and 2012(4) several
sport governing bodies convened international symposia to develop a consensus
statement and recommendations regarding the safety and health of athletes who
suffer concussion injuries. The agreement statements from this group provide a
concise and consistent definition of concussion, a description of clinical issues,
assessment tools, and return to play protocol. The consensus definition of sport-
related concussion is “a complex pathophysiological process affecting the brain,
induced by a traumatic biomechanical force.” (1-3) This definition is accompanied
by several common symptoms that are used in defining the severity of a
concussion.
Concussions are characterized by acute and long-term changes in the brain
resulting in a variety of physical (e.g. balance problems, loss of consciousness),
cognitive (e.g. slowed reaction time, memory loss, concentration difficulty), and
emotional (e.g. emotional lability) symptoms.(3, 5) These symptoms may present
immediately at the time of injury, or appear in the hours and days following the
injury.(6-10) The clinical symptoms used to diagnose concussion have been
described in detail(2, 3, 6, 9, 11-16). The acute symptoms that have been
21
validated as prognostic of injury progression and recovery include amnesia, loss
of consciousness, headache, dizziness, blurred vision, attention deficit, and
nausea.(14, 17, 18) Additional subjective symptoms have been described, but
have not been proven as prognostic features of concussion recovery.(18)
Sport-related concussions are of particular concern in young athletes.
Approximately 30 million children and adolescents in the United States
participate in sports every year.(33, 35) Surveillance in emergency departments
between 2001 and 2005 indicates that among patients examined for sport-related
concussion, a majority were pediatric patients. The highest rates were seen in
young athletes aged 10 to 19.(33) Given the physical and cognitive changes
occurring during puberty, adolescents are particularly vulnerable for concussion.
Injury surveillance and other research consistently find the highest rates of
adolescent sport-related concussion in football, wrestling, soccer, lacrosse and
basketball.(36-40)
While most cognitive detriments associated with concussion are the same
in children as in adults, the fact that the adolescent brain is actively maturing
may result in significant secondary consequences in children as their ability
perform day-to-day tasks impacts their educational and social attainment.(58)
Even in the absence of immediate presentation of clinical symptoms commonly
observed in adult concussion, injury to the adolescent brain occurs at a lower
energy threshold than in adults that manifest in the presentation in symptoms in
the post acute period.(59, 60) It is also suspected that despite normal
22
neuropsychological presentation following sport concussion, some children may
suffer a variety of behavioral and cognitive sequelae.(59, 61)
In addition to different injury mechanics, there is also evidence that
adolescent brains do not follow the same general recovery pattern as adults.
While adults generally resolve symptoms within the first week after concussion
injury, recent research indicates that young athletes may begin showing new
cognitive symptoms 7-10 days following injury.(33, 35) The adolescent brain may
be more sensitive to metabolic changes following injury than adult brains. This
imbalance may impact recovery time, partially explaining the delayed recovery
trajectory observed in adolescents compared to adults.(33, 50) History of
previous concussion also increases recovery time in adolescents. Subsequent
concussions demonstrate more severe immediate symptom presentation and are
associated with long-term neurological deficits.(33, 36, 51)
Understanding the recovery of adolescent athletes following concussion is
essential to minimize the consequences for the future health of the athletes.
Implementation of neurocognitive testing has improved the evaluation and
management of individual sport related concussion and has also allowed study of
the neurocognitive symptom presentation and progression following injury. An
increasing number of published studies present results from computerized
neurocognitive tests, allowing for examination of trends in differing populations
of athletes.
This section of the dissertation reviews the current literature related to
adolescent concussion injuries, summarizing the neurocognitive effects of
23
concussion at six or more days following injury in adolescent athletes. This
review evaluates the evidence for persisting symptoms and cognitive deficits in
adolescent athletes following concussion injuries.
2.2 Methods
A review of current literature addressing cognitive effects of concussion in
adolescents was accomplished employing the following terms: “cognitive effects”
or “neurocognitive effects” combined with “concussion,” “postconcussion” and/or
“mild traumatic brain injury” and “adolescent”. Searches were conducted in
PubMed and Ovid to identify studies published in English between January 1995
and May 2013. A two-step screening process was used to identify studies for
inclusion in the review. First, studies were reviewed at the abstract level and
restricted to those that reported neurocognitive test results for high school age
subjects following concussion or MTBI. A full text review of the remaining studies
restricted papers further and studies reporting results only from the immediate
period following injury (within 5 days) were excluded to focus on symptoms that
persist beyond the acute period. Papers were further excluded during the full text
review if they did not report neurocognitive outcomes or did not define
concussion exposure.
A standardized protocol was used to abstract information from the studies
identified for review on study design, study population, covariates, concussion
diagnosis and definition, outcome measurement, results, and relative study
quality. Use of quality scores in meta-analyses of observational studies is not
24
standard practice due to lack of demonstrated validity.(84) For this review,
assessments of study quality based on components of study design were noted
during the abstraction process, but quality scores were not integrated in the
pooled analyses. Results from neurocognitive tests were pooled where possible to
construct overall effects. This was not possible for all studies due to differences
in study populations, lack of standardized outcome testing, or incomplete data
presented in the paper. Where pooling was not possible, results were presented
as a systematic review with no compilation.
Several domains of cognition were measured by standard neurocognitive
testing tools. Eight of the included studies reported results from Immediate
Postconcussion Assessment and Cognitive Testing (ImPACT) assessment, a
computerized test of neurocognitive function commonly used to assess symptoms
following concussion. Four reported results from other neurocognitive testing
tools. Regardless of the test, these tools measure neurocognitive function in
several domains. Cognitive function consists of the common domains of
memory, attention, and executive function. Different aspects of these domains
are assessed by computerized neurocognitive testing tools. The results presented
in the studies reviewed here included assessment of function in the following
subdomains: Verbal, Visual, or Composite Memory; Processing Speed; Reaction
Time; and Total Symptom Score.
25
2.2.1 Statistical Methods
Studies reporting neurocognitive results from ImPACT were considered
for pooled analysis. Results were pooled where ImPACT results were reported for
the group with concussion during followup and for baseline (pre-injury) or a
control group. Pooled standardized mean differences (SMD) were calculated
using Cohen’s method and DerSimonian and Laird random effect weighting with
heterogeneity estimates taken from the inverse-variance fixed effect model.(85,
86) The random effects model allows for variation between studies and results
and provides a more conservative result than fixed effects models. Cornfield’s
confidence intervals were calculated.(87)
Heterogeneity was assessed through the I2 statistic and visualization of a
Galbraith radial plot. In the case of heterogeneity, sensitivity and influence
analyses were conducted by re-estimating the meta-analysis by omitting each
study in turn. An individual study may contribute excessive influence to the
analysis if the estimate of its omitted analysis falls outside the pooled result
confidence interval.
Publication bias was considered by visualizing a funnel plot and
performing the Beggs and Mazumdar adjusted rank correlation test (a statistical
analog to the funnel plot) and the Egger regression asymmetry test for
publication bias. Begg’s method and Egger’s tests assess the asymmetry of funnel
plots with correlation tests; Begg’s method uses rank correlation and Egger’s test
applies a regression to the standard normal deviate on precision.(88) All analyses
were performed using Stata/IC 11.2 for Mac (StataCorp, College Station, TX).
26
2.3 Results
After resolving duplicates, the search produced 152 unique publications.
Reference lists from each were also searched, yielding 9 additional publications.
Studies were excluded where concussion was not the exposure studied (n=101),
review papers or studies not designed to report neurocognitive results (n=33),
studies that did not report neurocognitive testing results 5 or more days following
concussion injury (n=14). The remaining twelve studies were included in the
review. Eight of the studies were cohort studies reporting results using ImPACT
testing. Three additional cohort studies and a cross sectional study reported
results from other neurocognitive tests. A schematic of the selection of
publications for review is presented in Figure 2.1.
The included studies are summarized in Table 2.1, presenting information
on demographics, methodological information, as well as primary findings.
Eleven of the studies were prospective or retrospective cohort studies and one
was cross sectional. A majority of the participants in all studies were male, and
the average age of participants was 15-17 years. Included studies represent results
for adolescent participants from across the United States and from Denmark. Six
of the studies reported baseline neurocognitive results for comparison, three
presented neurocognitive test results for a control group, and three offered no
comparison group. Collins et al(89) compared results between two concussed
groups – one with headache at the time of follow-up and one group without
headache at follow-up.
27
2.3.1 Studies reporting results from ImPACT neurocognitive testing
Eight studies reporting results from the ImPACT testing tools were
highlighted in Table 2.2. ImPACT results were presented for the following
domains: total symptom score, verbal memory, visual memory, composite
memory, processing speed, and reaction time.
2.3.1.1 Total Symptom Score Results
Table 2.2 provides information from studies reporting ImPACT total
symptom scores (N=7), with results ranging from 3.2(89) to 23.6(90). The
ImPACT total symptom score is a count of symptoms present at the time of
testing. Control and baseline results for studies presenting results range from
4.2(91) to 9.9(35). The “normal” range for total symptom scores on the ImPACT
test is 1-6 (49 to 76th percentile, as defined by ImPACT test creators)(92). Average
change from baseline or difference from control total symptom scores ranging
from a decrease of 3.3 points(35) (one study reported an improvement in
symptoms) 7 days following injury, to an increase of 7.9 points(93) 6-14 days
following injury (four studies(54, 89, 91, 93, 94) report a worsening of
symptoms). These average changes were summarized in Figure 2.2, a forest plot
showing the average change in each study and a quantitative summary measure
(SMD=0.52, 95% CI 0.05-0.99) Moser, et al(90) does not present baseline or
comparison group information, but presents a mean total symptom score of 23.6
(standard deviation 21.5) for subjects 8 to 30 days following injury. This result
was much greater than the normative range of 1 to 6.(92)
28
2.3.1.2 Verbal Memory Score Results
In the domain of verbal memory, four studies (54, 90, 93, 95) reported a
range of scores in subjects with concussion from 75.88(93) to 85(95). Average
baseline results were reported in two of the studies as 82.96(54) and 85.75(93).
The two studies(54, 93) reporting baseline results demonstrated a decrease in
verbal memory score 7 days following injury, although results from Covassin, et
al(54) were not statistically significant. Thomas, et al(95) and Moser et al(90),
did not report baseline results and offered no comparison group. The average
results for these studies fell within normal ranges defined by test creators. The
“average” range for males age 13 to 18 years is 80 to 92, for females age 13 to 18
the range is 84 to 93(92). The forest plot in Figure 2.3 was constructed to
illustrate the mean change in score for the two studies with baseline results and a
quantitative summary change (SMD= -0.66, 95% CI = -0.88 to -0.45)
2.3.1.3 Visual Memory Score Results
In four studies(54, 90, 93, 95), visual memory scores reported for subjects
with concussion range from 63.11(54) to 77(95) with two studies reporting
baseline scores of 71.25(54) and 74.04(93). There was a statistically significant
decrease in visual memory following injury in the two studies providing baseline
data. Thomas, et al(95) and Moser, et al(90), did not provide baseline data or a
comparison group. The results for these studies were with in the normal range
for visual memory test scores in adolescents. The “average” range for males age
16 to 18 years is 71-88, for males age 13 to 15 is 69 to 86, and for females age 13 to
29
18 is 70 to 88(92). In Figure 2.4, a forest plot presents the average change in
visual memory scores for the two studies that included baseline data as well as an
overall measure of change following injury (SMD= -0.5, 95% CI -0.71 to -0.29).
2.3.1.4 Composite Memory Score Results
Three studies(35, 89, 91) reported composite memory scores, rather than
individual verbal and visual memory scores. In subjects with concussion, results
ranged from 74.9(89) to 84.1(91) with baseline results reported in two of the
studies as 83.4(91) and 83.9(35). ImPACT does not provide average or normal
ranges for this score. Two studies report baseline composite memory scores, one
showing a decline on average in scores following injury(35), the other showing a
slight improvement on average in scores following injury(91). The third
study(89) reporting composite memory scores provided data on two groups with
concussion, one group with headache and the other without headache at follow-
up. The group with headache had statistically significant lower average
composite memory scores than the group without headache.(89)
2.3.1.5 Processing Speed Score Results
Processing speed in subjects with concussion was reported in six
studies(54, 89-91, 93, 95) with a range from 30.7(89) to 36.55(54). Processing
speed is measured by ImPACT as the number of tasks completed in a given time
with higher scores representing better performance. Baseline scores reported in
the included for processing speed ranged from 28.3(91) to 35.43(54). The
30
“average” range for males age 16 to 18 is 33.7-42.5, for males age 13 to 15 is 30.2
to 37.8, and for females age 13 to 18 is 32.8 to 42.3(92). In Figure 2.5, the
average change in processing speed scores was illustrated in a forest plot.
2.3.1.6 Reaction Time Score Results
Finally, reaction time in subjects with concussion ranged from 0.56(95) to
0.68(94) in seven studies(54, 89-91, 93-95). Control and baseline scores ranged
from 0.573(93) to 0.63(91). Reaction time is measured by ImPACT in seconds
with higher scores indicating poorer performance. Two studies presented
statistically significant results indicating a poorer reaction time following
injury.(93, 94) One study showed an increased reaction time that was not
statistically significant(54), and one study showed a decreased reaction time(91).
Thomas, et al(95) did not provide baseline data or a comparison group. Results
for this study group were within the normal range for reaction time in
adolescents. Moser, et al(90) did not provide baseline or comparison group data
for their study group. Results for this study were greater than the average range
for reaction time in adolescents. The “average” range for males age 16 to 18 year
is 0.50 to 0.58, for males age 13 to 15 is 0.53 to 0.60, and for females is 0.51 to
0.60.(92) Average change in reaction time is presented in a forest plot, Figure
2.6, with an overall measure of change (SMD=0.66, 95% CI -0.14 to 1.45)
31
2.3.1.7 ImPACT Scores Pooled Analysis Results
Table 2.3 presents the pooled SMD in each cognitive domain as well as for
each study included in the analysis. Total symptom score (pooled SMD= 0.524,
95% CI 0.054 to 0.993, p=0.029), verbal memory (pooled SMD=-0.662, 95% CI -
0.877 to -0.0445, p=0.006), and visual memory (pooled SMD=-0.500, 95% CI -
0.712 to -0.288, p<0.001) all resulted in statistically significant standardized
mean difference when pooled. Verbal memory and visual memory pooled
analyses were based on only two studies each. We observed no statistically
significant difference for standardized mean difference for processing speed
(pooled SMD=0.182, 95%CI -0.231 to 0.595, p=0.388) or reaction time (pooled
SMD=0.657, 95%CI -0.139 to 1.453, p=0.106).
Fixed effects models produced similar results to the random effects model,
however the more conservative estimates from the random effects model is
presented as the Galbraith plots and I2 statistics (TSS I2 = 85.6%, Verbal Memory
I2= 77.0%, Processing Speed I2= 77.6%, Reaction Time I2=92.9%) provided
evidence of significant heterogeneity among the studies. For the domain of visual
memory, the models appeared to have no heterogeneity (Visual Memory I2=
0.0%), however, only two studies were included in this analysis and the result is
unreliable. Publication bias was assessed through visualization of funnel plot
symmetry. Funnel plot asymmetry was assessed statistically using Begg’s method
and Egger’s test. However, with small numbers of studies included in these
analyses, the power of the test was low and the results are not reliable. Sensitivity
analysis was performed due to the indication of heterogeneity between the
32
studies. In analyses with four or more studies included, the influence of each
study on the SMD was determined by re-estimating the analysis while omitting
each study in turn. Each estimated SMD was still within the confidence interval
for the pooled SMD so the analysis was not modified to account for influence.
2.3.2 Studies reporting results from other neurocognitive tests
Teasdale, et al(96) reported results for men who suffered concussion
between the ages of 16 and 24 (n=700). Neurocognitive testing was the Broge
Prien Prove assessment tool administered by the draft board anywhere from 3 to
500 days following injury. Of this group, fourteen men were tested between
seven and thirty days following injury and eight were tested three to six days
following injury. Raw test scores were not presented, and results were
dichotomized as above or below the normal range defined by the Denmark draft
board. In the first week following concussion, injured men were more likely
(Relative Risk 2.45, 95%CI 1.23 – 4.90) to be classified as below normal range on
the neurocognitive battery.
Field, et al(97) used the Post Concussion Symptom Scale, Hopkins Verbal
Learning Test (HVLT), Brief Visual Spatial Memory Test-Revised (BVMT-R) and
Symbol Digit Modality Test to assess neurocognitive function following injury.
The study included 183 male high school varsity athletes and controls matched by
sport, age, grade point average, history of learning disability, and history of
previous concussion. In this population high school athletes had significant
verbal learning and memory impairment as measured by the HVLT compared to
33
controls remaining 7 days following injury. While a decline in visual memory
(measured by BVMT-R) compared to controls was noted, it was not statistically
significant (p=0.09).
Sim, et al (98) reported neurocognitive function based on the ANAM
testing tool. The prospective study included 419 athletes in Nebraska (14 with
concussion during the study period) who participated in the Nebraska
Concussion Study. The study participants were students in grades 9 through 12
in 9 public high schools. The study population consisted of mostly Caucasian
males with a mean age of 15.69 years. Compared to the control group, the
subjects with concussion had statistically significant deficits in reaction speed,
processing speed, and memory 2.5 days following injury. At approximately six
days following injury participants with concussion continued to show significant
memory impairments and score differences remained in processing speed and
reaction time, but were not statistically significant.
In a cross sectional study, Moser, et al(99) reported neurocognitive
function from the Repeatable Battery for the Assessment of Neuropsychology
Status (RBANS) on 35 athlete volunteers (14 with recent concussion) from a
highly academically competitive private school. The study population had a mean
age of 16.65 years and approximately 80% were male. Compared to a control
group with a history of zero or 1 concussions (not within the preceding 6
months), athletes with recent concussion (one week prior to assessment) had
poorer attention and memory scores.
34
2.4 Discussion
This review presents findings from studies that examined cognitive effects
six or more days following concussion in adolescent athletes. The review is
currently very relevant given the volume of research regarding the effects of
MTBI. The reviewed studies provide evidence that adolescent athletes experience
a range of neurocognitive symptoms six or more days following concussion
injury. Analysis of the published literature supported declines in the domains of
reaction speed, memory, and total symptom scores at the time of follow up
among adolescent athletes with sport-related concussion.
These studies support the current understanding of adolescent athlete
recovery from concussion injury. Another recent systematic review found that
younger adolescents had more severe neurocognitive impairment following
concussion prepared to college athletes and adults.(100) These results are also
reinforced by a recent report indicating that females, adolescents, and athletes
with 10 or fewer years of education demonstrated larger cognitive deficits than
males, adults, and athletes with more than 16 years of education in the first week
following injury.(101) While adults generally resolve symptoms within the first
week after injury, research indicates that young athletes may begin showing new
cognitive symptoms 7-10 days following injury.(33, 35) The adolescent brain may
be more sensitive to metabolic changes following injury than adult brains,
impacting recovery time and partially explaining the delayed recovery trajectory
observed in adolescents compared to adults.(33, 50) Adolescents with subsequent
35
concussions demonstrate more severe immediate symptom presentation and may
experience longer-term neurological deficits.(33, 36, 51)
Recovery trajectories following concussion injuries are highly individual
and difficult to predict. Severity and duration of neurocognitive and other
symptoms are influenced by individual characteristics, injury conditions, and
return to play progression.(71, 81, 102, 103) Deficits in different cognitive
domains may be the result of injury site, as different parts of the brain control
different cognitive function.(104, 105) On average, however, evidence is
increasing that adolescents will experience symptoms that persist beyond the
acute period of the injury.
This systematic review is complicated by several limitations. The body of
literature that the studies were pulled from is limited by factors such varying
definitions of concussion, and different measures of neurocognitive function.
While many studies report results from ImPACT, the current literature includes
results from several other neurocognitive test batteries. Publication bias is also of
concern, as the presented studies (no matter how systematically reviewed and
presented) may not present the full range of study results available on the
subject.
The meta-analyses of studies including ImPACT results are additionally
limited by factors that include small study populations, heterogeneity of results,
and residual confounding. With the exception of visual memory scores, analyses
indicate significant heterogeneity of reported results. The indication of no
heterogeneity in the visual memory results is likely an artifact of having only two
36
studies included in the test. Statistical and visual assessment of funnel plots to
assess the potential for publication bias are also questionable due to the small
number of studies included.
Heterogeneity between the studies could be explained in part by
differences in the study populations, and by the small number of studies included
in each test. These studies were conducted across the United States and include
subject with a wide range of demographic characteristics. Additionally,
documentation of ImPACT assessment protocols are not available, meaning that
the testing conditions may not be standardized across or within study
populations. Finally, the heterogeneity of results could also be explained in part
by the variability in time lapsed between injury and assessment.
Statistical tests do not indicate significant publication bias. However, the
analyses include a small number of observational studies reporting moderate
change in scores, leaving room for residual confounding. Residual confounding
in the meta-analysis could result from a number of potential factors that are not
included in study results including age, race, history of concussion, sport played,
severity of injury, and other factors.
Despite the limitations noted, results of this review and pooled analysis
support the current understanding of the trajectory of cognitive recovery
following concussion injuries in adolescent athletes. Many domains of cognition
remain impaired in adolescent athletes in the post acute period (>5 days after
injury). These findings support the need for further study of cognitive effects of
concussion in adolescents following the post acute period. These findings
37
support the importance of the research presented in this dissertation. Lingering
cognitive symptoms in adolescent athletes have the potential to impact academic
performance following concussion injuries.
38
2.5 Tables and Figures
Figure 2.1. Schematic illustrating the process of selecting publications for the systematic review.
Figure'1:'Schematic'illustrating'selection'of'publications'for'the'systematic'review.''
''Database'search'keywords:''''''''''''“concussion”'or'“postconcussion”'or'“mild'traumatic'brain'injury”'''''''''''“cognitive'effects”'or'“neurocognitive'effects”'''''''''''“adolescent”'Dates'searched:'January'1995'to'May'2013'
Number'of'unique'publications:'152'
Exclude'studies'''''''''Concussion'not'exposure'studied:'101'''''''''Review'papers:'9'''''''''Studies'not'reporting'neurocognitive'results:'24'''''''''Studies'not'reporting'results'>5'days'following'injury:'14'
Articles'for'full'review:'12'
Cohort'Studies'reporting'results'using'ImPACT'testing:'8'
Studies'reporting'results'from'other'neurocognitive'tests''''''Cohort'studies:'3''''''Cross'sectional:'1'
39
Tabl
e 2.
1. S
umm
ary
of in
clud
ed st
udy
dem
ogra
phic
s, m
etho
dolo
gy, a
nd p
rim
ary
findi
ngs.
Stud
y:'
N'
Follo
w-up'
Perio
d'Ag
e'rang
e'or''
mean'(sd)'
Gen
der''
(%'m
ale)'
Assess-
men
t'To
ol(s)'
Prim
ary'Find
ings'
Field,'2003'
183'
'7'days'
14'–'18'years'
100%'
HVLT,'
BVMT;
R,SDMT'
Verbal'learning'and'm
emory'impairment'(p<0.04)'and'visual'memory'
impairment'(p<0.09)'in'athletes'with'concussion'compared'to'controls'
Moser,'2002'
35'
7'days'
14'–'19'years'
80%'
RBANS'
Decreased'attention'(p=0.018)'and'm
emory'scores'(p;0.023)'in'athletes'
with'recent'concussion'compared'to'athletes'with'no'recent'concussion'
Sim
,'2008'
419'
4;7'days;'
8;11'days'
15.5'(1.09)'conc;'
15.69'(1.16)'no'
conc'
75%'
ANAM'
At'4;7'days:'significant'impairment'in'sim
ple'reaction'tim
e'(p=0.0001),'
information'processing'speed'(p=0.005)'and'delayed'm
emory'(p=0.006)'
in'concussed'athlete'compared'to'noninjured'control'and'their'own'
baseline.'
Teasdale,'1997'
700'
3;500'
days'
16;24'years'
100%'
Broge'
Prien'
Prove'
In'the'first'week'following'concussion'injured'm
en'were'm
ore'likely'to'
be'classified'below'normal'range'on'the'neurocognitive'assessment'
(Relative'Risk'2.45,'95%CI'1.23'–'4.90)'
Collins,'2003'
109'
5;10'days'
15.8'(1.2)'years'
84.5%'
ImPACT'
Athletes'with'concussion'who'present'with'headache'perform'worse'in'
total'symptom'score,'composite'm
emory,'and'reaction'tim
e'compared'
to'concussed'athletes'with'no'concussion'
Covassin,'2012'
72'
7'&'14'
days'
14'–'25'years'
71%'
ImPACT'
Compared'to'preinjury'results,'athletes'with'concussion'had'a'
statistically'significant'decline'in'the'domains'of'total'symptom'score,'
visual'memory,'and'reaction'tim
e'and'had'nonsignficant'decline'in'
verbal'memory'and'processing'speed.'
Lovell,'2003'
64'
7'days'
High'School'
94%'
ImPACT'
Compared'to'preinjury'results,'athletes'with'concussion'had'a'
statistically'significant'decline'in'composite'm
emory'scores.'
Lovell,'2004'
43'
6'days'
13'–'18'years'
81%'
ImPACT'
Compared'to'baseline,'athletes'with'concussion'had'statistically'
significant'decline'in'the'domains'of'total'symptom'score,'composite'
memory,'processing'speed,'and'reaction'tim
e.'
McClincy,'2006'
104'
7'days'
16.11'(2.22)'years'
81.5%'
ImPACT'
Compared'to'baseline,'athletes'with'concussion'had'statistically'
significant'decline'in'total'symptom'score,'verbal'memory,'visual'
memory,'and'reaction'tim
e.'
Maugans,'2012''
24'
14'days'
11'–'17'years'
75%'
ImPACT'
Compare'to'a'control'group,'athletes'with'concussion'had'poorer'scores'
on'total'symptoms'and'reaction'tim
e.'
Moser,'2012'
49'
8;30'days'
14'–'23'years'
67%'
ImPACT'
With'no'baseline'or'control'group'for'comparison,'athletes'with'
concussion'had'reaction'tim
e'scores'outside'the'“normal”'range.'
Thomas,'2011'
36'
6'days'
11'–'17'years'
78.3%'
ImPACT'
All'scores'were'within'the'“normal”'range.'No'baseline'or'control'group'
results'available'for'comparison.'
40
Tabl
e 2.
2. S
umm
ary
of in
clud
ed st
udie
s rep
ortin
g ne
uroc
ogni
tive
resu
lts fr
om Im
PACT
.
Stud
y:'
N'
Compa
rison
''#'da
ys'
after'
injury'
TSS'(sd)'
Verbal'
Mem
ory'(sd)'
Visual'M
emory'
(sd)'
Compo
site'
Mem
ory'
Score'(sd)'
Processing
'Spe
ed'
(sd)'
Reactio
n'Time'
(sd)'
Collins,'200
3''''''''Heada
che'
'''''''No'He
adache
'
' 36'
73'
Non
e''
7'' 21
.7'(2
2.3)*'§ '
3.2'(6.8)'
''
' 74.9(16.2)*'§ '
82.4(10.7)'
' 30.7(7.5)'
33.3(8.7)'
' 0.64
(0.09)*'§ '
0.57
(0.09)'
Covassin,'201
2''''''''Post;injury'
'''''''Baseline'
72'
Baseline'
7'' 11
.18(14
.69)*'§ '
6.29
(7.23)'
' 77.61(16
.02)
'§'
82.96(10
.38)'
' 63.11(15
.85)*'§ '
71.25(14
.49)'
'' 36
.55(9.4)
'§'
35.43(6.68
)'
' 0.60
9(0.14
9)'§'
0.59
2(0.07
5)'
Lovell,'200
3''''''''Post;injury'
'''''''Baseline'
64'
Baseline'
7'' 6.6(13
.9)*'
9.9(12
.9)'
''
' 80.2(13.1)*'§ '
83.9(8.6)'
''
Lovell,'200
4''''''''Post;injury'
'''''''Baseline'
43'
Baseline'
6'' 8.7(13
.1)*
'§'
4.2(8.6)'
' '' '
' 84.1(8.8)*'
83.4(9.4)'
' 32.8(8.5)*
'§'
28.3(5.3)'
' 0.59
(0.09)*'
0.63
(0.07)'
McClincy,'200
6''''''''Post;injury'
'''''''Baseline'
10 4'Ba
seline'
7'' 13
.08(15
.55)*'§ '
5.14
(7.87)'
' 75.88(13
.79)*'§ '
85.75(8.59
)'
' 66.96(15
.92)*'§ '
74.04(13
.82)'
'' 33
.99(8.1)'
35.05(6.9)'
' 0.63
5(0.12
1)*'§ '
0.57
3(0.07
6)'
Mau
gans,'201
2''
'''''''Post;injury'
'''''''Con
trol'Group
'
' 12'
12'
Control'
Group
'14
'' 11
.17(4.41
)*'§'
4.08
(1.26)'
''
''
' 0.68
(0.03)*'§ '
0.59
(0.02)'
Moser,'201
2'49
'Non
e'8;30
'23
.6(21.5)'
82.6(11.5)'
69.7(7.9)'
'33
.8(7.4)'
0.63
4(0.09
)'Th
omas,'201
1'36
'Non
e'6'
'85
'77
''
36.11'
0.56
'Im
PACT
'“no
rmal”'range1'
'''''''M
ales'(1
3;18
y)'
'''''''Fem
ales'(1
3;18
y)'
''
'1;6'
' 80;92'
84;93'
' 69;88'
70;88'
' '' 30
.2;42.5'
32.8;42.3'
' 0.50
;0.60'
0.51
;0.60'
*Sta
tistic
ally
sign
ifica
nt d
iffer
ence
bet
wee
n gr
oups
(p<0
.05)
§ R
esul
ts in
dica
te a
dec
line
in p
erfo
rman
ce fo
llow
ing
conc
ussi
on o
r com
pare
d to
con
trol
gro
up
1 Nor
mal
rang
es fo
r Im
PACT
test
in u
ninj
ured
ado
lesc
ents
are
pro
vide
d fo
r com
pari
son.
41
40
Table 2.3. Standardized mean differences for ImPACT cognitive domains. Cognitive)Domain) N) SMD) 95%)CI) Weight) p7value) I7squared)Total&Symptom&Score&&&&&&&&Covassin,&2012&&&&&&&&Lovell,&2003&&&&&&&&Lovell,&2004&&&&&&&&McClincy,&2006&&&&&&&&Maugans,&2012&
&72&64&43&104&12&
0.524&0.422&A0.246&0.406&0.644&2.186&
(0.054A0.993)&(0.092A0.753)&(A0.0594A0.102)&(A0.021A0.833)&(0.365A0.923)&(1.158A3.215)&
&22.42&22.15&20.86&23.16&11.41&
0.029& 85.6%&
Verbal&Memory&&&&&&&&Covassin,&2012&&&&&&&&McClincy,&2006&
&72&104&
A0.662&A0.396&A0.859&
(A0.877A&A0.4447)&(A0.726A&A0.066)&(A1.143A&A0.575)&
&48.29&51.71&
0.006& 77.0%&
Visual&Memory&&&&&&&&Covassin,&2012&&&&&&&&McClincy,&2006&
&72&104&
A0.500&A0.536&A0.475&
(A0.712A&A0.288)&(A0.869A&A0.203)&(A0.751A&A0.199)&
&40.72&59.28&
<0.001& 0.0%&
Processing&Speed&&&&&&&&Covassin,&2012&&&&&&&&Lovell,&2004&&&&&&&&McClincy,&2006&
&72&43&104&
0.182&0.137&0.635&A0.141&
(A0.231A0.595)&(A0.190A0.464)&(0.202A1.069)&(A0.413A0.131)&
&34.12&29.36&36.52&
0.388& 77.6%&
Reaction&Time&&&&&&&&Covassin,&2012&&&&&&&&Lovell,&2004&&&&&&&&McClincy,&2006&&&&&&&&Maugans,&2012&
&72&43&104&12&
0.657&0.144&0.496&0.614&3.530&
(A0.139A1.453)&(A0.183A0.471)&(A0.925A&A0.067)&(0.335A0.892)&(2.215A4.845)&
&28.07&27.14&28.45&16.33&
0.106& 92.9%&
Note:&Pooled&standardized&mean&differences&(SMD)&were&calculated&using&Cohen’s&method&and&DerSimonian&and&Laird&random&effect&weighting&with&heterogeneity&estimates&taken&from&the&inverseAvariance&fixed&effect&model.&
42
40
Figure 2.2. ImPACT Total Symptom Score Forest Plot. Note: Pooled standardized mean differences for ImPACT Total Symptom Score calculated using Cohen’s method and DerSimonian and Laird random effect weighting with heterogeneity estimates taken from the inverse-variance fixed effect model.
NOTE: Weights are from random effects analysis
0.52 (0.05, 0.99)
0.42 (0.09, 0.75)
-0.25 (-0.59, 0.10)
SMD (95% CI)
2.19 (1.16, 3.21)
0.41 (-0.02, 0.83)
0.64 (0.37, 0.92)
100.00
22.42
22.15
%
Weight
11.41
20.86
23.16
Covassin, 2012
Lovell, 2003
Lovell, 2004
McClincy, 2006
Maugans, 2012
Overall (I-squared = 85.6%, p=0.029)
SMD (95% CI) % Weight
0-3.21 0 3.21
43
40
Figure 2.3. ImPACT Verbal Memory Score Forest Plot. Note: Pooled standardized mean differences for ImPACT Verbal Memory Score calculated using Cohen’s method and DerSimonian and Laird random effect weighting with heterogeneity estimates taken from the inverse-variance fixed effect model.
Figure 2.4. ImPACT Visual Memory Score Forest Plot.
-0.66 (-0.88, -0.45)
SMD (95% CI)
-0.40 (-0.73, -0.07)
-0.86 (-1.14, -0.57)
100.00
Weight
42.60
%
57.40
Covassin, 2012
McClincy, 2006
Overall (I-squared=77.0%, p=0.006)
NOTE: Weights are from random effects analysis
SMD (95% CI) % Weight
0-1.14 0 1.14
-0.50 (-0.71, -0.29)
SMD (95% CI)
-0.47 (-0.75, -0.20)
-0.54 (-0.87, -0.20)
100.00
Weight
59.28
%
40.72Covassin, 2012
McClincy, 2006
Overall (I-squared=0.0%, p=<0.001)
NOTE: Weights are from random effects analysis
SMD (95% CI) % Weight
0-.869 0 .869
44
40
Note: Pooled standardized mean differences for ImPACT Visual Memory Score calculated using Cohen’s method and DerSimonian and Laird random effect weighting with heterogeneity estimates taken from the inverse-variance fixed effect model.
Figure 2.5. ImPACT Processing Speed Score Forest Plot. Note: Pooled standardized mean differences for ImPACT Processing Speed Score calculated using Cohen’s method and DerSimonian and Laird random effect weighting with heterogeneity estimates taken from the inverse-variance fixed effect model.
NOTE: Weights are from random effects analysis
0.18 (-0.23, 0.60)
0.64 (0.20, 1.07)
SMD (95% CI)
0.14 (-0.19, 0.46)
-0.14 (-0.41, 0.13)
100.00
29.36
Weight
34.12
36.52
%
Covassin, 2012
Lovell, 2004
McClincy, 2006
Overall (I-squared=77.6%, p=0.388)
SMD (95% CI) % Weight
0-1.07 0 1.07
Processing Speed
45
40
Figure 2.6. ImPACT Reaction Time Score Forest Plot. Note: Pooled standardized mean differences for ImPACT Reaction Time Score calculated using Cohen’s method and DerSimonian and Laird random effect weighting with heterogeneity estimates taken from the inverse-variance fixed effect model.
NOTE: Weights are from random effects analysis
0.66 (-0.14, 1.45)
-0.50 (-0.93, -0.07)
SMD (95% CI)
0.61 (0.34, 0.89)
0.14 (-0.18, 0.47)
3.53 (2.22, 4.84)
100.00
27.14
Weight
28.45
28.07
16.33
%
Covassin, 2012
Lovell, 2004
McClincy, 2006
Maugans, 2012
Overall (I-squared=92.9%, p=0.106)
SMD (95% CI) % Weight
0-4.84 0 4.84
Reaction Time
46
CHAPTER 3: RELATIONSHIP BETWEEN CONCUSSION INJURY OR LOWER LEG INJURY AND ACADEMIC PERFORMANCE IN HIGH
SCHOOL ATHLETES
47
Abstract
Background. Research indicates that adolescents may have prolonged or
new neurocognitive symptoms that appear beyond the post-acute period
following concussion injury. The aim of this study was to estimate the incidence
of injury and distribution of academic variables (including school attendance) in
a population of high school athletes with concussion injury and lower leg injury.
Study Design. A retrospective cohort of athletes participating in school
sports in a large suburban school district during the 2009-2010 academic year
was analyzed. A group of athletes with sport-related concussion recorded in the
electronic medical record and a comparison group who experienced ankle or leg
injuries during the study period were included in analyses of injury incidence
rates.
Methods. Incidence of injury during the study period was calculated as
the number of injuries per 10,000 athletic exposures. Rate ratios were calculated
comparing the rates of injury in concussion and ankle/leg injury and between
male and female athletes in comparable sports. Pearson correlation, chi-square,
and t-test statistics were used to compare injury severity in athletes with
concussion and ankle/leg injury. Odds ratios were calculated for odds of injury by
grade level, severity of injury by gender, and change in school attendance.
48
Results. A total of 494 concussion and 472 ankle/leg injuries were
recorded during the study period. Football accounted for the greatest proportion
of concussion (36%) and ankle/leg (23%) injuries as well as the highest rates of
concussion (5.19/10,000AE, 95%CI 4.42-5.96). Football (RR=1.69, 95%CI 1.32-
2.18) and lacrosse (RR=1.95 95%CI 1.30-2.96) had significantly higher rates of
concussion compared to ankle/leg injury. There was an increased odds of
concussion injury for athletes in grade 9 (OR=2.2, 95% CI 1.5-3.2) or grade 10
(OR=2.1, 95%CI 1.5-3.1), than grade 12 when compared to athletes with ankle/leg
injuries. We observed an increase in the number of days absent from school for
athletes with concussion (OR=1.7, 95% CI 1.3-2.2). The odds of severe
concussion was increased in female athletes compared to male athletes (OR=2.4,
95% CI 1.4-4.3).
Conclusions. Concussion injury rates in this population follow patterns
similar to other study populations with contact sports, female, and younger
athletes demonstrating higher rates of concussion. Athletes with concussion
injuries had greater odds of increased absence from school compared to athletes
with ankle or leg injuries. As school attendance is highly correlated with
academic performance, an increase in school absence may put athletes with
concussion at risk for poor academic achievement.
49
3.1 Introduction
Concussive head impacts are a natural part of most athletic competition.
Even in sports that do not permit player contact, there is incidental contact with
other players, equipment, or the playing surface that may result in injury. Sport-
related concussions, a subgroup of mild traumatic brain injury (mtbi), are
accompanied by a constellation of poorly defined short and long-term physical
and cognitive consequences for the athlete.
Concussions are characterized by acute and long-term changes in the brain
resulting in a variety of physical (e.g. balance problems, loss of consciousness),
cognitive (e.g. slowed reaction time, memory loss, concentration difficulty), and
emotional (e.g. emotional lability) symptoms.(3, 5) These symptoms may present
immediately at the time of injury, or appear in the hours and days following the
injury.(6-10) Although the clinical symptoms used to diagnose concussion have
been described in detail(2, 3, 6, 9, 11-16).
Sport-related concussions are of particular concern in young athletes.
Approximately 30 million children and adolescents in the United States
participate in sports every year.(33, 35) Surveillance in emergency departments
between 2001 and 2005 indicates that among patients examined for sport-related
concussion, a majority were pediatric patients. The highest rates were seen in
young athletes aged 10 to 19.(33) Given the physical and cognitive changes
occurring during puberty, adolescents are particularly vulnerable for concussion.
Injury surveillance and other research consistently find the highest rates of
50
adolescent sport-related concussion in football, wrestling, soccer, and
basketball.(36-40)
While most cognitive detriments associated with concussion are the same
in children as in adults, the fact that the adolescent brain is actively maturing
may result in significant secondary consequences in children as their ability
perform day-to-day tasks impacts their educational and social attainment.(58)
Even in the absence of immediate presentation of clinical symptoms commonly
observed in adult concussion, injury to the adolescent brain occurs at a lower
energy threshold than in adults that manifest in the presentation in symptoms in
the post acute period.(59, 60)
There is also evidence that adolescent brains do not follow the same
general recovery pattern as adults. While adults generally resolve symptoms
within the first week after concussion injury, recent research indicates that young
athletes may begin showing new cognitive symptoms 7-10 days following
injury.(33, 35) The adolescent brain may be more sensitive to metabolic changes
following injury than adult brains. This imbalance may impact recovery time,
partially explaining the delayed recovery trajectory observed in adolescents
compared to adults.(33, 50) History of previous concussion also increases
recovery time in adolescents. Subsequent concussions demonstrate more severe
immediate symptom presentation and are associated with long-term neurological
deficits.(33, 36, 51)
Understanding the recovery of adolescent athletes following concussion is
essential to minimize the consequences for the future health of the athletes.
51
Implementation of neurocognitive testing has improved the evaluation and
management of individual sport related concussion and has also allowed study of
the neurocognitive symptom presentation and progression following injury. An
increasingly number of published studies present results from computerized
neurocognitive tests, allowing for examination of trends in differing populations
of athletes.
In this section of the dissertation, the distribution and rates of concussion
injuries and ankle or leg injuries in a high school athlete population from a large
school district were examined. The analyses presented here provide descriptive
statistics for sport-related concussion in high school athletes during the 2009-
2010 academic year and their social and educational characteristics in
comparison to a population of athletes with lower extremity injury.
3.2 Methods
3.2.1 Subjects.
This retrospective study includes 966 participants that were grade 9
through 12 during the 2009-2010 academic year, from one school district
consisting of 25 schools. The school district serves a county with approximately
one million residents characterized by a an average median household income
nearly twice the national average.(106) The school district enrolls over 175,000
students annually and those students perform at a higher level than other
students in their state and in the US on standardized tests.(107) [Note: the name
of the school district has been omitted from this text and blinded in the
references as a condition of IRB approval of this project.] The school district
52
employs certified athletic trainers to support all athletic programs. The athletic
trainers monitor all athlete participation in sport practice and competition,
documenting all injuries in an electronic medical record (EMR).
The study population for this research was restricted to students who were
participating in scholastic athletics and had records in the electronic medical
records maintained by district employed athletic trainers for injuries documented
during the 2009-2010 academic year. Inclusion criteria were students with
documented sport-related concussion or ankle/lower leg injury during the study
period. For inclusion in the study, athlete records were required to include
academic information for the study period and data available on injury recovery
and return to play. Academic records were not available for students who were
not enrolled in the study district prior to the quarter of their injury. For example,
an athlete in the ninth grade with an injury during the fall quarter would not have
prior academic data available (n=211). These subjects were included in analyses
that did not consider prior academic data. During this period, concussion
assessment and rehabilitation protocols were consistent across all schools in the
district. Additional information regarding time lost from sport and academic time
lost was collected. Data regarding injury type, severity, time lost, sport, and date
were pulled from the district EMR. Academic data including absence from
school, and grade level were extracted from district academic records. The
institutional review board of Drexel University and the Research Screening
Committee of the study school district approved the study prior to obtaining
student records.
53
3.2.2 Concussion Definition.
Athletes were categorized as exposed for analyses if they had a
documented concussion during the academic year 2009-2010. District employed
athletic trainers monitor all athlete participation in sports practice and
competition, and document all athletic injuries in an electronic medical record
(EMR) system for students participating in 27 club and varsity sports. Athletic
trainers at official scholastic games and practices diagnose concussion according
to a standardized protocol. The assessment tool used has been adapted from the
Standardized Assessment of Concussion (SAC).(15, 16) The SAC is administered
immediately following injury by a certified athletic trainer and assesses
orientation, immediate and delayed memory, concentration, exertional
maneuvers, and a brief neurological screening. If the injured athlete scores below
a previously measured baseline, the athlete is not returned to play and is referred
for follow-up and rehabilitation with athletic trainers. Injury assessment and
rehabilitation information is recorded in the EMR by the athletic trainer on a
daily basis. Annual in-service sessions provide ongoing training in concussion
recognition and management. The athletic trainers have frequent contact with
athletes, increasing recognition of changes in behavior associated with
concussion. Athletes with complicated presentation of symptoms are referred to
physicians for care. Reevaluation and rehabilitation services are provided at the
school’s athletic training facility by district trained athletic trainers.
54
3.2.3 Comparison Group Definition
Athletes without concussion during the study period who had sport-
related ankle or lower leg injuries documented in the district EMR were included
in analyses as a comparison group. To be included in the comparison group,
athletes must have a documented ankle or lower leg injury during the 2009-2010
academic year, no concussion during the study period, and academic information
available for the study period.
3.2.4 Variables
Variables used in analyses include athletic exposure, time lost from sport
(injury severity), absence from school, grade level, gender, and sport played at
the time of injury.
To calculate injury rates for the population, the number of times the
athlete is at risk for injury (either during athletic training sessions or
competitions) was included as the denominator measure of exposure to represent
person-time. Athletic exposure (AE) was calculated as the number of athletes
participating in a sport multiplied by the number of sport training sessions or
competitions for a season. For example, a sport with 50 athletes competing in
100 practices or competitions would result in denominator of 5000 AE in the
injury rate calculation.
For these analyses, average annual athletic exposure was based on the
participation data for 2010-2011 and 2011-2012 academic years, as participation
55
data for the 2009-2010 study period was not available. Participation numbers do
not fluctuate substantially from year to year.
Absence from school was included in study data as the raw number of days
that a student was not in school during an academic period. For stratified
analyses, a change in school attendance was calculated as the difference in the
number of days a student was absent from school during the academic period of
injury compared to the academic period prior to injury. The change in school
attendance was categorized as “less” if the athlete had fewer days absent in the
period of injury compared to the period prior to injury, “same” if the number of
days absent was the same, and “more” if the athlete had more days absent from
school in the period of injury compare to the period prior to injury.
Time lost from sport was calculated as the number of days elapsed from
the date of the injury to the date the athlete was returned to play. Time lost from
sport was used as a measure of injury severity, with more severe injuries
requiring a longer period of time before the athlete returns to play. This measure
is a good indication of time required for recovery from injury and is used in these
analyses to represent the severity of the injury. Gender was not reported directly
in study data and was inferred in a subset of the study population from the sport
played where gender was indicated in the name of the sport. For example,
athletes coded as playing “Girls Soccer” were inferred to be female, while those
playing “Boys Soccer” were inferred to be male.
Missing data were identified for each variable. School absences were
missing in the quarters prior to injury for students who were not enrolled in the
56
district high school system prior to the quarter of their injury. These subjects
were included in calculation of incidence rates, but not in analyses of change in
absence from school.
3.2.5 Statistical Analysis.
Injury rates were calculated for each sport and stratified by gender when
possible. Injury rates per 10,000 athletic exposures were calculated as a measure
of the incidence of injury during the study period. Using athletic exposures as the
denominator allows for comparison of injury rates between populations. Rate
ratios were calculated to compare the rates of injury in concussion and ankle/leg
injury populations as well as between male and female athletes in comparable
sports. Pearson chi-square statistics and student t-tests were used to compare
injury severity (measured by time lost from sport) in athletes with concussion to
athletes with ankle/leg injury. Odds ratios were calculated for odds of injury
severity by gender and odds of increased absence from school by injury type. All
analyses were performed with Stata/IC 11.2 for Mac. Results for comparative
statistics were considered statistically significant at a p value <0.05.
3.3 Results
A total of 494 concussion injuries and 472 ankle/leg injuries were
recorded during the study period, inclusive of injuries without prior academic
data available. Information about injury severity and academic information was
not available for all subjects. 86 athletes with concussion and 47 athletes with
57
ankle or leg injury did not have prior academic data available for analyses. These
subjects were included in calculation of injury rates, but not in calculations that
require academic variables prior to injury. Ten athletes with concussion and five
athletes with ankle or leg injury did not have information on time lost from sport.
These subjects were excluded from calculation of injury severity.
Football accounted for the greatest proportion of concussion injuries
(36%) and ankle/leg injuries (23%) and the highest concussion injury rate
(5.19/10,000AE 95%CI 4.42-5.96). Football (Rate Ratio=1.69 95%CI 1.32-2.18)
and lacrosse (Rate Ratio=1.95 95%CI 1.30-2.96) both resulted in statistically
significant higher rates of concussion than ankle/leg injuries. Basketball (Rate
Ratio=0.56 95%CI 0.33-0.93), cross country/track (Rate Ratio=0.20 95%CI
0.10-0.37), and soccer (Rate Ratio=0.69 95%CI 0.48-1.00) resulted in
statistically significant lower rates of concussion injuries compared to ankle/leg
injuries (Table 3.1). Among concussion injuries, football, baseball/softball,
basketball, and wrestling had a higher proportion of injuries that resulted in
more than 14 days of time lost than injuries that resulted in less than 10 days of
time lost (Table 3.2).
Table 3.3 provides information about the distribution of injuries by type
and grade level. Ankle and leg injuries were evenly distributed by grade level
with approximately 25% of injuries occurring in athletes at each grade level. In
the group with concussion injuries, however, a higher proportion of injuries
occurred in younger athletes. There were increased odds of concussion injury in
58
athletes in grade 9 (OR=2.2, p<0.001, 95% CI 1.5-3.2) or grade 10 (OR=2.1,
p<0.001, 95%CI 1.5-3.1), when compared to athletes in grade 12.
The mean (standard deviation) number of days absent from school in the
quarter prior to injury was 1.57 (1.93) for athletes with concussion and 1.48 (1.82)
for athletes with ankle or leg injury. In the quarter of injury, the mean (standard
deviation) number of absences for athletes with concussion was 2.29 (2.84,
p<0.001) and 1.54 (2.00, p=0.53) for athletes with ankle or leg injury. Athletes
with concussion injury had an odds of missing more days of school in the
academic period of injury that is 1.7 times that of having the same or fewer days
missed in the academic period of the injury (OR=1.7, p<0.011, 95% CI 1.3-2.2).
For 164 athletes with concussion (34% of athletes with concussion),
gender could be inferred from the sport played (67 male athletes, 97 female
athletes). In this subpopulation of athletes female athletes account for a greater
proportion of concussion in each sport with the exception of lacrosse. Table 3.5
details the distribution of concussion injuries by gender and sport. Rate ratios for
sports where gender could be inferred illustrate that females have a greater rate
of concussion injuries than male athletes (basketball Rate Ratio=3.72(95% CI
1.43-11.38), cross country/track Rate Ratio=2.69(95% CI 0.72-12.22), lacrosse
Rate Ratio=1.10(95% CI 0.67-1.78), soccer Rate Ratio=3.11(95% CI 1.65-6.20).
Female athletes were more likely to have more severe concussions than male
athletes participating in comparable sports (Table 3.6). Female athletes had a
mean time lost from sport of 25.79 days (standard deviation 1.81) with
concussion and 20.21 days (standard deviation 6.08) with ankle or leg injury.
59
Males had a mean time lost form sport of 21.85 days (standard deviation 1.55)
with concussion and 14.5 days (standard deviation 1.48) with ankle or leg injury.
While the difference in mean days lost from sport was not statistically significant
for female athletes compared to male athletes, odds ratios comparing days lost
from sport indicated that female athletes had higher odds of concussion injuries
with time lost from sport of more than 14 days than injuries with time lost of less
than 10 days compared to male athletes with concussion playing comparable
sports (OR=2.4, p=0.002, 95% CI 1.4-4.3). This association is not evident when
athletes with ankle and leg injuries were compared (OR=1.25, p=0.89, 95%CI
0.81-1.93), there was not a statistically significant difference in the number of
days lost from sport based on gender.
3.4 Discussion These descriptive analyses confirm that the study data are consistent with
other study populations and provide information about subpopulations, injury
severity, and comparative results for the two populations of athletic injury.
During this study period, 494 concussion injuries and 473 ankle or leg
injuries were recorded. Many of the results presented here are consistent with
other studies examining concussion injuries and recovery in high school athletes.
First, football (36% of concussion injuries, injury rate=5.19/10,000AE, 95%CI
4.42-5.96) and lacrosse (15% of concussion injuries, injury rate= 4.77/10,000AE,
95%CI 3.68-5.89) account for a large proportion of concussion injuries in this
population. These result align with results from a 10 year prospective study
60
(between academic years 1997-1998 and 2007-2008) of concussion rates in the
same study population that found high rates of concussion in athletes
participating in these collision sports (39). This trend is true in other study
populations, as well.(37, 45) In a sample of 100 US high schools, Gessel et al(37)
studied injury rates in 9 sports during the 2005-2006 academic year. Their study
documented an overall concussion rate of 2.3 injuries per 10,000 AE with the
highest rate occurring in football participants (4.7 concussion injuries per 10,000
AE). (37) Marar et al calculated an injury rate of 2.5 concussions per 10,000AE
for a national sample of high school injuries in 20 sports, with the highest
concussion rates resulted from participation in football (6.4), boys’ hockey (5.4)
and boys’ lacrosse (4.0) and a higher rate of concussion in girls than boys
(RR=1.7, 95%CI 1.4-20).(45)
Additionally, concussion severity results in this study are consistent with
data from other populations, as well.(45, 108, 109) These studies presented data
indicating that concussions in football, baseball, basketball, and wrestling are
more likely to have prolonged recovery compared to concussions occurring in
other sports. In this study, football (mean 21.68, sd 2.27, p=0.004),
baseball/softball (mean 31.79, sd 6.72, p=0.037), basketball (mean 18.12, sd 3.04,
p=0.005), and wrestling (mean 27.6, sd 2.76, p=0.030) had a higher proportion
of concussion injuries that resulted in more than 14 days of time lost from sport
than concussion injuries that resulted in less than 10 days of time lost.
Finally, in the subpopulation where gender could be inferred from the
sport played, female athletes experienced rates of concussion approximately 3
61
times higher than male athletes in each sport except lacrosse. It should be noted
that lacrosse rules are different for females and males, female athletes are not
allowed as much direct contact as male athletes, making lacrosse less comparable
between genders than sports that do not have rule modification by gender. The
rate ratio for female:male concussion rates in basketball (Rate Ratio=3.72,
p=0.0014), cross country/track (Rate Ratio=2.69, p=0.54), and soccer (Rate
Ratio=3.11, p=0.001) indicated higher rates of concussions experienced by female
athletes. Female athletes had higher odds of more severe concussion than male
athletes participating in equivalent sports (OR=2.4, 95%CI 1.36-4.34). These
trends are not evident when comparing athletes with ankle and leg injuries
(OR=1.25, 95%CI 0.81-1.93).
In this study population, a significantly higher proportion of concussion
injuries in younger athletes compared to ankle or leg injuries was observed. Odds
of concussion injury were higher for athletes in grade 9 (OR=2.2, 95%CI 1.5-3.2)
or grade 10 (OR=2.1, 95%CI 1.5-3.2) than in grade 12 when compared to athletes
with ankle and leg injuries. This result is consistent with younger athletes being
more susceptible to concussion injuries (33, 36), but could also be explained by
younger athletes being more willing to openly report injury symptoms than older
athletes. It is also possible that athletes at highest risk for concussion injury that
experience concussion at younger age are no longer participating in sport at later
grades.
Athletes in this study population with concussion injury were nearly twice
as likely to have a significant increase in the number of days absent from school
62
than athletes with ankle or leg injuries. This result is of particular interest in the
context of this dissertation as absence from school is closely tied to academic
performance. Students who are absent from the classroom miss instructional
time and are more likely to perform poorly on classroom and standardized
assessment.(110)
The data from this population are frequently used in other athletic injury
research and while the dataset used for these analyses was limited, the overall
quality of the data is high with few missing records and a history of population
data for comparison.
Limitations of this study include the limited data available on the study
population and variables that could have contributed to the understanding of the
data presented. For example, access to gender and calendar age were not
available in the analytic data set. These variables were inferred from data
available, which provide reasonable proxy measures. Where possible, gender was
inferred from the sport played. Calendar age was not provided in the analytic
data set. Instead, academic grade level is used to represent developmental age.
Most state athletics regulatory associations have strict age eligibility rules,
ensuring that athletes participating in high school athletics are under age
nineteen.
Diagnosis of concussion is dependent on accurate self-report of symptoms
by the injured athlete and the athletic trainer correctly recognizing symptoms.
Underreporting of symptoms by athletes is well documented (39, 91, 111, 112) and
necessarily means that not all cases are reported. Concussions that are diagnosed
63
and recorded may not be representative of all actual concussive incidents. These
analyses were limited to subjects with injuries that were diagnosed and recorded
in the EMR of the study school district, presenting a potential for selection bias.
Athletes whose concussions were recorded in the EMR were likely more severe
cases. Therefore, these results may not apply to athletes with less severe
concussion injuries. Additionally, athletes with concussions that were not
recorded due to unreported symptoms may differ from the subjects with a
comparable level of symptoms who do report. These data do not capture the
athletes whose injuries occurred outside of the district athletic programs, unless
follow-up with a district athletic trainer was documented. There are likely other
injuries that occurred during recreational activity or prior to enrollment in the
district. These incidents of injury are not identified in the EMR. However, the
study population represents the athletes accessible for intervention and follow-up
by district personnel.
External validity of these results is of concern due to the fact that the study
school district has committed resources and attention to athletic programs that
are not present in other schools. The injury experience of athletes in this study
may not be reflective of the experience of athletes in other school systems where
athletic programs do not include medical support.
These analyses provide a foundational understanding of the distribution of
injuries and injury rates experienced by athletes in the study population. The
data from this population are frequently used in other athletic injury research
and while the data did not provide specific measure of some potential
64
confounders, the overall quality of the data is high with few missing records and a
history of injury rates in this population for comparison.
We found a higher proportion of concussion injuries in younger athletes
compared to ankle or leg injuries and a higher rate and more severe concussions
in female athletes compared to male athletes playing comparable sports. We
also found a statistically significant increase in the number of days absent from
school in the period of injury for athletes with concussion injury. This finding is
of particular interest as absence from school is closely tied to academic
performance(110, 113) and this population of athletes is at risk for decreased
performance on classroom and standardized assessment due to missed
instructional time.
Missed instructional time due to sport-related injury is of interest to
educational and medical professionals treating high school athletes. Current
concussion rehabilitation protocols recommend cognitive rest following injury to
ensure full recovery. In high school athletes, prolonged cognitive rest may result
in absence from school which impacts instructional time and may compound
effects on academic performance. Further study of the association between
academic performance and athletic concussion injury is important to understand
this risk. Individual concussion rehabilitation plans may need to consider a
balance between prolonged cognitive rest and return to classroom instruction to
mitigate effects on academic performance.
65
3.
5 Ta
bles
and
Fig
ures
Ta
ble
3.1.
Dis
trib
utio
n of
inju
ry b
y sp
ort a
nd in
jury
type
.
Sport&
Average&&
Annu
al&
Athletic&
Expo
sures&
CONCU
SSION&IN
JURIES&
ANKLE/LEG&IN
JURIES&
&
N&(%
)&Injury&ra
te&per&10,00
0&AE
&(95%
&CI)&
N&(%
)&
Injury&ra
te&per&
10,000
&AE&
(95%
&CI)&
Rate&Ratio&
(95%
&CI)&
Rate&ra
tio&
pJvalue&
&Ba
seba
ll/Softba
ll,11
7,48
4,19
,(4),
1.62
,(0.8992.34
),,18
,(4),
1.53
,(0.8392.24
),1.06
,(0.5292.13
),0.44
,Ba
sketba
ll,,
137,31
6,25
,(5),
1.82
,(1.1192.53
),45
,(10),
3.28
,(2.3294.24
),0.56
,(0.3390.93
),0.00
9,Ch
eer,a
nd,Dan
ce,
99,715
,36
,(8),
3.61
,(2.4394.79
),27
,(6),
2.71
,(1.6993.73
),1.33
,(0.7992.28
),0.13
,Cross,C
ountry,and
,Track
1,51
1,51
0,12
,(3),
0.23
,(0.1090.37
),61
,(14),
1.19
,(0.8991.49
),0.20
,(0.1090.37
),<0
.000
1,Field,Ho
ckey,
92,167
,12
,(3),
1.30
,(0.5792.04
),11
,(2),
1.19
,(0.4991.90
),1.10
,(0.4492.73
),0.42
,Footba
ll,*,
335,41
8,17
4,(36),
5.19
,(4.4295.96
),10
3,(23),
3.07
,(2.4893.66
),1.69
,(1.3292.18
),<0
.000
1,Lacrosse,*,
155,18
1,74
,(15),
4.77
,(3.6895.89
),38
,(9),
2.45
,(1.6793.23
),1.95
,(1.3092.96
),0.00
3,Soccer,
146,08
6,53
,(11),
3.63
,(2.6594.61
),77
,(17),
5.27
,(4.0996.45
),0.69
,(0.4891.00
),0.01
8,Wrestling,
66,776
,30
,(6),
4.49
,(2.8996.10
),21
,(5),
3.14
,(1.8094.49
),1.43
,(0.7992.63
),0.11
,Other,sp
orts
2,26
6,80
5,45
,(9),
1.69
,(1.1992.18
),44
,(10),
1.65
,(1.1692.14
),1.02
,(0.6691.59
),0.46
,TO
TAL,
1,92
7,73
8,48
0,(100
),2.49
,(1.7893.01
),44
5,(100
),2.3,(1.6593.07
),,
,P9values,were,calculated
,from
,Pearson
,chi9squ
are,diffe
rences,in,injury,ra
te,by,sport.,,,
Average,an
nual,athletic,exposures,are,based
,on,20
109201
1,an
d,20
119201
3,pa
rticipation,da
ta,fo
r,the
,stud
y,distric
t.,“*”,Indicates,a
,statistically,highe
r,rate,of,con
cussion,than
,ankle/le
g,injury,
1,Cross,C
ountry,and
,Track,includ
es,cross,cou
ntry,,ind
oor,track,,and
,outdo
or,track,
2 “Other”,sports,includ
e,tenn
is,,sw
imming,,volleyball,,go
lf,,crew,,and
,injurie
s,tha
t,hap
pen,ou
tside,of,sc
hool,sp
orts,
66
Table 3.2. Severity of concussion injuries (as represented by days lost from sport) by sport.
Sport& Mean&(sd)&
Days&lost&from&sport&N(%)&
<10&days& 10914&days& >14&days& p9value1&
Baseball/Softball*- 31.79(6.72)- 1-(5)- 3-(16)-- 15-(79)- 0.037-
!!!!!!!!!!Male!!!!!!!!!!!Female!
19!32.5(7.07)!
0!1!
0!3!
1!14!
-
Basketball-- 18.12-(3.04)- 11-(44)- 6-(24)- 8-(32)- 0.005-
!!!!!!!!!!Male!!!!!!!!!!!Female-
12.3!(3.78)!19.95!(3.77)!
4!7!
1!5!
1!7!
!
Cheer-and-Dance*- 25.91(4.26)- 8-(22)- 5-(14)- 23-(64)- 0.0063-
Cross-Country-and-Track*- 40.5-(9.05)- 1-(8)- 3-(25)- 8-(67)- 0.022-
!!!!!!!!!!Male!!!!!!!!!!!Female-
20.5(10.22)!50.5(11.30)!
1!0!
2!1!
1!7!
!
Field-Hockey- 26.92(6.69)- 3-(25)- 2-(17)- 7-(58)- 0.97-
Football*- 21.68(2.27)- 48-(28)- 52-(30)- 74-(42)- 0.004-
Lacrosse*- 22.61(2.04)- 14-(19)- 17-(23)- 43-(58)- 0.037-
!!!!!!!!!!Male!!!!!!!!!!!Female-
21.74(2.55)!23.93(3.41)!
10!4!
11!6!
21!18!
!
Soccer- 21.35-(3.06)- 10(19)- 16-(30)- 27-(51)- 0.072-
!!!!!!!!!!Male!!!!!!!!!!!Female!
17.36!(3.29)!22.82!(4.01)!
4!6!
3!13!
7!19!
-
Wrestling*- 27.6(2.76)- 3-(10)- 3-(10)- 24-(80)- 0.030-
Other-sports- 31.79(6.72)- 13-(22)- 7-(12)- 38-(66)- 0.49-
!!!!!!!!!!Male!!!!!!!!!!!Female-
10!22(4.20)!
0!2!
1!3!
0!6!
-
POvalues-were-calculated-from-Pearson-chiOsquared-differences-in-the-proportions-comparing-the-proportion-
of-athletes-with-the-most-severe-injuries-(>14-days-lost-from-sport)-with-the-least-severe-injuries-(<10-days-
lost-from-sport)--
“*”-indicates-sports-with-a-statistically-significant-higher-proportion-of-severe-injuries-than-minor-injuries-
(p<0.05)
67
Figure 3.1 Severity of concussion injury by sport. Note: Stacked bars represent the proportion of days lost from sport in each severity category: less than 10 days, 10-14 days, and more than 14 days. “*” indicates sports with a statistically significant(p<0.05) higher proportion of severe injuries (>14 days lost) than minor injuries (<10 days lost). Table 3.3. Distribution of injury by grade level. Grade&Level& Concussion&
Injuries&N&(%)&
Ankle/Leg&Injuries&N&(%)&
OR&(95%&CI)& p9value&
9-(Freshman)-*- 157-(32)- 118-(25)- 2.2-(1.5O3.2)- <0.001-
10-(Sophomore)*- 162-(33)- 125-(26)- 2.1-(1.5O3.2)- <0.001-
11-(Junior)- 105-(21)- 114-(24)- 1.5-(1.0O2.3)- 0.05-
12-(Senior)- 70-(14)- 116-(25)- REFERENCE- -
TOTAL- 494-(100)- 473-(100)- - -
POvalues-were-calculated-from-Pearson-chiOsquare-differences-in-proportion-of-each-type-of-injury-by-grade-level.--
“*”-indicates-with-a-statistically-significant-higher-proportion-of-concussion-injuries-(p<0.05).-
1-
11-
8-
1-
3- 48-14- 10-
3-
13-3-
6-
5-
3-
2-
52-
17-16-
3-
7-
15-
8-
23- 8-7-
74-
43-27-
24-
38-
0%-
10%-
20%-
30%-
40%-
50%-
60%-
70%-
80%-
90%-
100%-
<10-days- 10O14-days- >14-days-
68
Table 3.4. Change in school attendance by injury type.
Change&in&school&attendance& Concussion&Injury&
Ankle/Leg&Injury&
Odds&Ratio&(95%&CI)&
p9value&
- Mean(sd)& Mean(sd)&-
-
Absence-prior-to-injury- 1.57(1.93)- 1.48(1.82)- 0.46-
Absence-in-period-of-injury- 2.29-(2.84)- 1.54(2.00)- <0.001-
- &N(%)&
&N(%)&
--
Less-absence-than-prior-to-injury- 149-(31)- 156-(33)-REFERENCE-
<0.001-Same-number-of-absences- 117-(24)- 157-(34)-
More-absences-than-prior-to-
injury-
222-(45)- 153-(33)- 1.7-(1.3O2.2)-
O-Change-in-school-attendance-was-calculated-as-the-difference-in-the-number-of-days-absent-from-school-in-the-
academic-period-of-the-injury-compared-to-the-academic-period-prior-to-injury.---
O-Odds-ratio-compares-the-odds-of-an-increase-in-school-absence-compared-to-no-change-or-fewer-absences-in-
athletes-with-concussion-versus-athletes-with-ankle-or-leg-injuries.
Figure 3.2 Severity of concussion injury by sport and gender. Note: Stacked bars represent the proportion of days lost from sport in each severity category: less than 10 days, 10-14 days, and more than 14 days.
1-
4-
7-1- 10-
4-
4-
6- 2-3-
1-
5- 2-
1-
11-
6-
3-
13-
1-
3-
1-
14-
1-
7-1-
7-
21-
18-
7- 19- 6-
0%-
10%-
20%-
30%-
40%-
50%-
60%-
70%-
80%-
90%-
100%-
<10-days- 10O14-days- >14-days-
69
Table 3.5. Distribution of concussion injuries by gender and sport. & Male&Athletes& Female&Athletes& & &Sport& Average&
annual&athletic&exposures&
Number&of&concussion&injuries&(%)&&
Average&annual&athletic&exposures&
Number&of&concussion&injuries&(%)&&
Rate&Ratio&(95%CI)&
p9value&
Basketball-*- 74,158- 6-(9)- 63,158- 19-(20)- 3.72-
(1.43O11.38)-
0.0014-
Cross-Country-and-
Track-
293,469- 4-(6)- 218,043- 8-(8)- 2.69-
(0.72O12.22)-
0.054-
Lacrosse- 93,591- 43-(64)- 61,591- 31-(32)- 1.10-
(0.67O1.78)-
0.35-
Soccer-*- 77,067- 14-(21)- 69,019- 39-(40)- 3.11-
(1.65O6.20)-
0.001-
TOTAL- - 67-(100)- - 97-(100)- - -
POvalues-were-calculated-from-Pearson-chiOsquare-comparison-of-rates-of-injuries-by-gender-for-each-sport-where-
gender-could-be-inferred.-
Table 3.6. Odds of injury severity by gender. &Days&lost&from&sport&
Male&Athletes&N&(%)&
Female&Athletes&N&(%)&
Odds&Ratio&(95%&CI)&
p9value&
Concussion-Injury- - - - -
!!Mean!(sd)! 21.85(1.55)! 25.79!(1.81)! ! 0.10!<10-days- 67-(27)- 20-(16)- REFERENCE- -
10O14-days- 70-(29)- 31-(24)- 1.4-(0.77O2.85)- 0.24-
>14-days- 106-(44)- 77-(60)- 2.4-(1.36O4.34)- 0.002-
Ankle/Leg-Injury- - - - -
!!!Mean!(sd)! 14.50!(1.48)! 20.21!(6.08)! ! 0.30!<10-days- 141-(57)- 101-(55)- REFERENCE- -
10O14-days- 41-(17)- 24-(13)- 0.82-(0.46O1.44)- 0.48-
>14-days- 65-(26)- 58-(32)- 1.25-(0.81O1.93)- 0.89-
Odds-ratios-compare-female-athletes-to-male-athletes-and-the-least-severe-injuries-(<10-days-lost-from-sport)-as-
the-reference.-
70
CHAPTER 4: LONGITUDINAL ANALYSIS OF ACADEMIC PERFORMANCE IN HIGH SCHOOL STUDENTS WITH ATHLETIC
INJURIES
71
Abstract
Background. Understanding the recovery of adolescent athletes
following concussion, including the effects on academic performance, is essential
to minimize the consequences for the future health of the athletes. The aim of this
study is to use longitudinal statistical methods to characterize the impact of
sport-related concussion on academic performance in high school athletes.
Study Design. A retrospective cohort of athletes participating in school
sports in a large suburban school district during the 2009-2010 academic year
was analyzed. A group of athletes with sport-related concussion recorded in the
electronic medical record and unexposed comparison group who experienced
ankle or knee injuries during the study period were included in analyses.
Academic grades and absence from school in the two academic periods prior to
injury, the academic period of the injury, and the two academic periods following
the injury were analyzed. Subgroup analyses were conducted to explore the
effects in younger athletes and gender groups.
Methods. Students t-tests were conducted to test for a change in mean
GPA from the academic period prior to injury and the academic period of injury,
effect sizes were calculated based on Cohen’s method. Repeated measures
MANOVA were conducted to test for differences between groups on mean GPA in
the academic period prior to injury and the academic period of the injury.
Growth curve models were constructed to visualize the mean change in academic
72
performance over time. Random coefficient models were used to explore and
explain average trends in mean GPA and individual differences in mean GPA.
Results. 494 concussion injuries and 473 ankle/leg injuries were
recorded during the study period. While there was not an overall main effect
noted in this population, there is an indication that younger athletes with more
severe injury are at risk for poor academic outcomes following injury. Athletes
with both concussion injury and ankle or leg injury show a decline in academic
performance in the academic period of injury and those that follow. Both groups
also demonstrate an increase in the number of days absent from school in the
period of injury compared to the period prior to injury. Analyses of a subgroup of
younger athletes (grades 9 and 10) with more severe injuries there was a 16%
greater decline in academic performance for the group with concussion injuries
compared to those with ankle or leg injury (coeff=-0.18, p=0.017, 95% CI=-0.32
to -0.03).
Conclusions. High school athletes with sport-related concussion are at
risk for decline in academic performance following injury. Athletes with both
concussion and ankle or leg injury experience an increase in the number of days
absent from school, which can negatively impact academic achievement.
Research regarding development of targeted instructional interventions that
provide opportunities for academic support as well as cognitive rest in an effort to
get students back to pre-injury performance levels is warranted.
73
4.1 Introduction
Sport related injuries have consequences beyond the physical
manifestations of the injuries. School-aged athletes with injuries have higher
rates of school absence than students without injury(114-116), which has the
potential to impact academic performance. Symptoms associated with
concussion such as concentration and memory difficulties, may further impact
student performance in school.
Compared to adults, adolescents are at increased risk for concussion with
greater severity and prolonged recovery from symptoms.(3, 71, 117) While most
cognitive detriments associated with concussion are the same in children as in
adults, the fact that the adolescent brain is actively maturing may result in
significant secondary consequences in children as their ability perform day-to-
day tasks impacts their educational and social attainment.(58)
A review of published studies examining cognitive effects 6 to 14 days
following concussion in adolescents, provides evidence that adolescent athletes
experience a range of neurocognitive symptoms 6 or more days following
concussion injury.(118) Athletes in the reviewed studies demonstrated declines
in the domains of reaction speed, memory, and total symptom scores at the time
of follow up.(118)
These studies support the current understanding of adolescent athlete
recovery from concussion injury. While adults generally resolve symptoms
within the first week after injury, recent research indicates that young athletes
may begin showing new cognitive symptoms 7-10 days following injury.(33, 35)
74
The adolescent brain may be more sensitive to metabolic changes following
injury than adult brains, impacting recovery time and partially explaining the
delayed recovery trajectory observed in adolescents compared to adults.(33, 50)
Adolescents with subsequent concussions demonstrate more severe immediate
symptom presentation and may experience longer-term neurological deficits.(33,
36, 51)
Recovery trajectories following concussion injuries are highly individual
and difficult to predict. Severity and duration of neurocognitive and other
symptoms are influenced by individual characteristics, injury conditions, and
return to play progression. Deficits in different cognitive domains may be the
result of injury site, as different parts of the brain control different cognitive
function. On average, however, evidence is increasing that adolescents will
experience symptoms that persist beyond the acute period of the injury.
The aim of this section of the dissertation was to explore the longitudinal
trends of academic performance in high school athletes with concussion injury
and ankle/leg injury using random effects models to produce effect estimates.
4.2 Methods
4.2.1 Study Subjects.
This retrospective study includes participants that were grade 9 through 12
from one school district consisting of 27 schools. The school district serves a
county with approximately one million residents characterized by a higher than
average median household income (approximately twice the national median
75
household income).(106) The school district enrolls over 175,000 students
annually and those students perform at a higher level than other students in their
state and in the US on standardized tests. The school district employs certified
athletic trainers to support all athletic programs. The athletic trainers monitor all
athlete participation in sport practice and competition, documenting all injuries
in an electronic medical record (EMR).
The study population for this research was restricted to students who were
participating in scholastic athletics and had records in the electronic medical
records maintained by district employed athletic trainers for injuries documented
during the 2009-2010 academic year. Inclusion criteria were students with
documented sport-related concussion or ankle/lower leg injury. During this
period, concussion assessment and rehabilitation protocols were consistent
across all schools in the district. Additional information regarding time lost form
sport and academic time lost was collected. The institutional review board of
Drexel University and the Research Screening Committee of the study school
district approved the study prior to obtaining student records.
4.2.2 Concussion Definition.
Athletes were categorized as exposed for analyses if they had a
documented concussion during the study period. For inclusion in the study,
athlete records were required to include academic information for the study
period and data available on injury recovery and return to play. Sport-related
concussions were diagnosed by district employed athletic trainers using a
76
standardized protocol. The athletic trainers monitor all athlete participation in
sports practice and competition, and document all athletic injuries in an
electronic medical record (EMR) system for students participating in 27 club and
varsity sports. Athletic trainers at official scholastic games and practices diagnose
concussion according to a standardized protocol. This assessment tool has been
adapted from the Standardized Assessment of Concussion (SAC).(15, 16) The SAC
is administered immediately following injury by a certified athletic trainer and
assesses orientation, immediate and delayed memory, concentration, exertional
maneuvers, and a brief neurological screening. If the injured athlete scores below
a previously measured baseline, the athlete is not returned to play and is referred
for follow up and rehabilitation with athletic trainers. Injury assessment and
rehabilitation information is recorded in the EMR by the athletic trainer on a
daily basis. Annual in-service sessions provide ongoing training in concussion
recognition and management. The athletic trainers have frequent contact with
athletes, increasing recognition of changes in behavior associated with
concussion. Athletes with complicated presentation of symptoms are referred to
physicians for care. Reevaluation and rehabilitation services are provided at the
school’s athletic training facility by district trained athletic trainers.
4.2.3 Comparison Group Definition
Athletes without concussion during the study period who had ankle or
lower leg injuries were included in analyses as a control group. To be included in
the control group, athletes must have a documented ankle or lower leg injury
77
during the study period, no concussion during the study period, academic
information available for the study period, and data on recovery and return to
play.
4.2.4 Outcome Definition.
Subject specific and overall grades were obtained as measures of academic
performance. The study school district has a common grading scale across all
schools. Grades for two total quarters are considered in t-test and MANOVA
analyses – the quarter prior to injury and the quarter of the injury. Grades for
five academic periods (two prior to injury, the quarter of the injury, and two
quarters following injury) are analyzed using longitudinal methods.
4.2.5 Variables
Variables used in analyses include time lost from sport, absence from
school, grade level, gender, and sport played at the time of injury.
Absence from school was included in study data as the raw number of days
that a student was not in school during an academic period. For analyses, a
change in school attendance is calculated as the difference in the number of days
a student was not in school during the academic period of injury compared to the
academic period prior to injury. The change in school attendance was
categorized as “less” if the athlete had fewer days absent in the period of injury
compared to the period prior to injury, “same” if the number of days absent was
78
the same, and “more” if the athlete had more days absent from school in the
period of injury compare to the period prior to injury.
Time lost from sport was calculated as the number of days elapsed from
the date of the injury to the date the athlete was returned to play. Time lost from
sport is used as a measure of injury severity, with more severe injuries requiring a
longer period of time before the athlete returns to play. Gender is not reported
directly in study data and was inferred in a subset of the study population from
the sport played where gender is indicated in the name of the sport. For example,
athletes coded as playing “Girls Soccer” are inferred to be female, while those
playing “Boys Soccer” are inferred to be male.
In an effort to identify athletes with injuries early in the quarter to allow
time for an effect on academic performance, analyses were restricted to athletes
with injuries in the first half of the quarter. This restriction on the study
population did not change the result.
4.2.6 Statistical Analysis.
Unpaired student’s t-tests were conducted to test for a change in mean
GPA from the academic period prior to injury and the academic period during
which the injury occurred. Data are presented for athletes with concussion and
with ankle or leg injuries, further subdivided by severity into varying periods of
time lost for sport. Effect sizes were calculated based on Cohen’s method, which
estimates the magnitude of the effect by describing the mean difference between
79
groups in terms of standard deviations(119). To calculate Cohen’s d, the mean
difference is divided by the pooled standard deviation for the samples.
Repeated measures MANOVA were conducted to test for differences
between groups on mean GPA in the academic period prior to injury and the
academic period of the injury. Type I errors were assessed through simulation,
due to violations of the normality assumptions for the ANOVA test. Simulations
were performed, computing sample sizes and standard deviations, assuming
equal means. The simulation produces a p value for 5,000 ANOVA and F-test
simulations. If this number deviates from the reported p-value for the study data
analyses, type I error is more likely.
Growth curve models were constructed to visual the mean change in
academic performance over time. Random coefficient models were used to
explore and explain average trends in mean GPA and individual differences in
mean GPA by allowing random variation in subject-specific relationships with
piecewise splines where both the level of response and the change is allowed to
vary between groups after controlling for covariates. Polynomial terms were
included to correct for nonlinear trends in the outcome variable. Model fit was
assessed using likelihood ratio tests.
Analyses were restricted to younger athletes (academic years 9 and 10)
with more severe injuries (time lost from sport greater than or equal to 10 days)
for subgroup analyses, to test the a prior hypotheses predicting greater potential
for effect in these groups.
80
All analyses were performed with Stata/IC 11.2 for Mac. Results were
considered statistically significant at a p value < 0.05. Analyses were restricted to
several subpopulations to assess the effects in specific groups. Interactions
between exposure, time, and grade level were considered to test a prior
hypotheses.
4.3 Results
A total of 494 concussion injuries and 473 ankle and leg injuries were
recorded during the study period. Of this population, 275 were in grade 9 at the
time of injury, 287 in grade 10, 219 in grade 11, and 186 in grade 12. Table 4.1
provides the distribution of each injury type in the population. Information about
injury severity and academic data were not available for all subjects.
4.3.1 Change in Mean GPA with Injury
Table 4. 2 presents results of student’s t-test exploring the change in mean
GPA from the academic period prior to injury to the academic period of injury in
athletes with concussion injury and athletes with ankle or leg injury.
Athletes with concussion injuries had a mean difference in GPA of -0.011 points
on a 4.0 point scale (sd=0.45, p=0.62) and athletes with ankle or leg injuries had
a mean difference in GPA of -0.035 (sd=0.44, p=0.10). (Table 4.2) A t-test
comparing these mean difference indicated no statistical difference (p=0.43).
In both groups, the decline in academic performance was greater and
statistically significant when the analyses were restricted to athletes with the
81
most significant injury (as measured by time lost from sport greater than 14
days). In this subpopulation, athletes with concussion had a mean difference in
GPA of -0.090 (d=0.14, sd=0.46, p=0.004) and athletes with ankle or leg injuries
had a mean difference in GPA of -.095 (d=0.15, sd=0.39, p=0.010).
Similar subanalyses were conducted for English, Math, and Science
grades, comparing grades in these subjects prior to injury to grades in the
academic period of the injury. No statistically significant differences were found
in the change in these grades before and after injury between the two injury
groups.
A repeated measures MANOVA test of the main effects of injury type,
grade level, and injury severity (as measured by time lost from sport) on the
mean GPA in the academic period prior to injury and the academic period of the
injury shows no main or interaction effects for injury type (p=0.20). Results are
presented in Table 4.3. Grade level (p=0.08), injury severity (p=0.03), and
increased absence from school (p<0.001) were significantly associated with
change in mean GPA. Simulations to assess type I error due to violations of the
normality assumption produced p-values that were not significantly different
from the study data, making type I error in these analyses unlikely.
4.3.2 Longitudinal change in mean GPA
Growth curve models were constructed to observe the mean change in
GPA over time. Figure 4.1 illustrates the mean change in GPA over time in
athletes with concussion injury and athletes with ankle or leg injury. Both groups
82
experienced a decline in mean GPA at the time of injury and in the academic
periods following injury.
Random effects analyses showed that GPA and absences vary both within
and between subjects (standard deviation both between and within subjects >0).
Grade level, injury type, and time lost from sport also varied between subjects
(standard deviation >0). Due to this variation, random coefficient models were
constructed allowing both the slope and intercept of the model to vary.
Likelihood ratio tests indicated that allowing the slope to vary does not improve
the regression model. Likelihood ratio tests comparing the model allowing the
intercept (individual mean GPA) to vary by subject resulted in a significantly
improved model (p<0.001) compared to a fixed effects model, therefore a
random intercept model was used for the analyses presented here.
Random intercept models indicated a decline in academic performance as
reported by mean GPA following both concussion and ankle/leg injury when the
model is controlled for injury severity, absence from school, and grade level.
After controlling for the number of days absent from school and severity of
injury, athletes with concussion experienced a 16% (coeff= -0.18, p=0.017, 95%
CI= -0.32 to -0.03) greater decline in average grades over the observation period
compared to athletes with ankle/leg injuries for injuries occurring in grades 9
and 10.
83
4.4 Discussion
In the analyses presented here, the academic performance effects of a
concussion injury were compared to other athletic injury. Compared to athletes
with ankle/leg injuries, high school students with sport-related concussion did
not suffer a significant decline in academic performance. Athletes in both injury
groups experienced a decline in mean GPA during the quarter of injury compared
to the quarter prior to injury. When analyses were restricted to younger athletes
(grades 9 or 10) or those with more severe injuries (time lost from sport greater
than 10 days), athletes with concussion injury demonstrated a greater decline in
average grades over the observation period than athletes with ankle or leg injury.
These analyses were limited by several factors. First, the dependent
variable was not measured in a standardized manner. The grade point average in
a given quarter was not necessarily based on the same classes as in a different
quarter. Variation in coursework and instructors may have an effect on
individual learning. Growth patterns in academic achievement are also
associated with decreasing cognitive capacity for new knowledge as a child
ages.(30, 57, 120) Curriculum and complexity of instruction also increase with
succession through academic grades which can affect student achievement.(120)
Additionally, the study sample was drawn from a single school district that is not
generalizable to other populations. The study school district has high student
achievement when compared to other school districts and students have access to
resources that may not be available in other districts.
84
The random coefficient model is limited by several factors. First, analysis
was limited by the assumption of a linear trend in the outcome variable. In this
case the trend trajectory was not linear, and a polynomial trajectory correction
term was included in the model. This complicates interpretation of the model
effect estimates, but capture the patterns of change over time and allow for
identification of differences in trends between groups.
Second, the random coefficient model is unable to account for
measurement error in the outcome variable as would be possible in a structural
equation model. This limitation is outweighed, however, by the flexibility of the
random coefficient model when modeling time across outcome measurements.
For example – each academic period results in a recorded grade that is based on
a variable number of graded elements over variable number of instructional days.
These data would be problematic for analysis in a structural equation model, but
do not violate any assumptions of the random coefficient model.
In the study population analyzed, there does not appear to be a significant
negative effect on academic performance following sport-related concussion in
older students or in students with mild concussion when compared to athletes
with ankle or leg injury. This finding does not rule out the possibility that
individual injury may have academic consequences for an athlete. In the school
district considered, athletes have access to care from certified athletic trainers to
aid in rehabilitation, and the school district as a whole has a greater awareness of
concussion injuries than other school districts that do not have access to
resources and training regarding concussion injury rehabilitation. This increased
85
awareness may be associated with increased accommodation and academic
support following injuries that mitigates the academic effects that are of concern.
There appears to be a significant decline in academic performance in
younger athletes (grades 9 or 10) with severe concussion injuries compared to
athletes with similarly severe ankle or leg injury (measured by time lost from
sport). This finding is consistent with the understanding that younger
adolescents are more susceptible to concussion injury due to the rapid
neurocognitive development occurring during puberty. All adolescent athletes
should be closely monitored following concussion injury, but particular attention
should be paid to the neurocognitive rehabilitation and academic return of
younger athletes. Additional cognitive rest or accommodation in the classroom
may be necessary to prevent negative effects on academic performance measures
in these athletes.
86
4.5 Tables and Figures Table 4.1. Distribution of injury in the study population. & Concussion&
Injuries&N&(%)&
Ankle/Leg&Injuries&N(%)&
p9value4&
Grade-Level- 494- 473- -
9-(Freshman)-- 157-(32)- 118-(25)- <0.001-
10-(Sophomore)- 162-(33)- 125-(26)-
11-(Junior)- 105-(21)- 114-(24)-
12-(Senior)- 70-(14)- 116-(25)-
Gender1-
371- 430- -
----Male- 243-(65)- 247-(58)-0.02-
----Female- 128-(35)- 183-(42)-
Severity2-
493- 430- -
----<10-days-lost- 112-(23)- 242-(56)- <0.001-
----10O14-days-lost- 114-(23)- 65-(15)-
---->14-days-lost- 267-(54)- 123-(29)-
Absence3- 488- 466- -
----Same-or-fewer-days- 266-(55)- 313-(67)-<0.001-
----More-days- 222-(45)- 153-(33)-1-Gender-was-inferred-from-the-sport-played-as-it-was-not-directly-described-in-the-study-data.---
2-Injury-severity-was-by-the-number-of-days-lost-from-sport.---3-Absence-from-school-was-calculated-as-the-difference-in-the-number-of-days-the-student-was-
not-in-attendance-during-the-academic-period-of-the-injury-compared-to-the-academic-period-
prior-to-injury.--4-POvalues-were-calculated-from-Pearson-chiOsquare-differences-in-proportion-of-each-type-of-
injury.-
87
Table 4.2. Comparison of mean GPA prior to injury to academic period of the injury. & N& Mean&
difference&Standard&Deviation&
p9value& Effect&Size&
Mean-GPA-for-academic-period-of-injury-compared-to-period-prior-to-injury-
--Concussion-Injury- 426- O0.011- 0.45- 0.62- 0.02-
-----<10-days-lost- 93- O0.034- 0.33- 0.18- 0.06-
-----10O14-days-lost- 96- 0.011- 0.47- 0.82- 0.02-
----->14-days-lost- 219- O0.31- 0.46- 0.004- 0.14-
--Ankle/Leg-Injury- 408- O0.035- 0.44- 0.10- 0.05-
----------<10-days-lost- 238- O0.032- 0.49- 0.43- 0.04-
----------10O14-days-lost- 67- 0.045- 0.37- 0.76- 0.02-
---------->14-days-lost- 120- O0.095- 0.39- 0.010- 0.15-
Overall-pOvalue-- - - - 0.43- -
Effect-sizes-are-calculated-based-on-Cohen’s-d,-pOvalues-test-the-null-hypothesis-that-there-is-no-
change-in-mean-GPA,-and-the-overall-pOvalue-tests-the-null-hypothesis-that-the-mean-difference-in-
GPA-is-the-same-in-athletes-with-concussion-and-athletes-with-ankle-or-leg-injury.-
Table 4.3. Effects of injury type and severity on mean GPA. & Wilks’&
Lambda&df& F& p9value& Effect&Size&
Model- 0.94- 8- 3.17- <0.001- -
--Injury-type- 0.99- 1- 1.62- 0.20- 0.08-
--Grade-Level- 0.99- 3- 1.86- 0.08- 0.06-
--Severity- 0.98- 3- 3.36- 0.03- 0.01-
--Absence- 0.98- 1- 7.6- <0.001- 0.11-
----1-Injury-type-O-concussion-vs-ankle/leg-injury-
-----2Grade-level-in-school-(9-through-12)-
-----3Severity-O-measured-by-time-lost-from-sport-
-----4Absence-–-increase-in-absence-from-school-compared-to-academic-period-prior-to-injury-
MANOVA-test-of-the-main-and-interaction-effects-of-injury-type,-grade-level,-and-injury-severity-(as-measured-by-
time-lost-from-sport)-on-the-repeated-measure-of-mean-grade-point-average-in-the-academic-period-prior-to-injury-
and-the-academic-period-of-injury.-
88
Table 4.4. Results of random intercept models assessing the academic impact of concussion based on severity and grade level. & Coefficient& Coefficient& 95%&CI& Model&& & p9value& Lower& Upper& p9value&Concussion-injury- O0.068- 0.127- O0.15- O0.29- 0.008-
Model-is-controlled-for-injury-severity,-absence-from-school,-and-grade-level-
-
Severity-
<10-days-lost- O0.044- 0.605- O0.21- 0.122- 0.21-
10O14-days-lost- O0.23- 0.015- O0.41- O0.05- 0.016-
>14-days-lost- O0.046- 0.48- O0.18- 0.08- 0.50-
Models-are-controlled-for-injury-type,-absence-from-school,-and-grade-level-
Severity-is-reflected-in-the-models-as-a-measure-of-time-lost-from-sport. -
Grade-Level-
9-(Freshman)- O0.17- 0.049- O0.34- O0.001- 0.039-
10-(Sophomore)- O0.10- 0.215- O0.27- 0.06- 0.016-
11-(Junior)- 0.018- 0.828- O0.14- 0.17- 0.54-
12-(Senior)- 0.032- 0.756----- O0.17- 0.23- 0.82-
Models-are-controlled-for-injury-type,-absence-from-school,-and-injury-severity-
Table 4.5. Results from random intercept model restricted to athletes in lower grades (grade 9 or grade 10) assessing the academic impact of concussion based on severity of injury. & Coefficient& Coefficient& 95%&CI& Model&& & p9value& Lower& Upper& p9value&Severity- - - - - -
<10-days-lost- O0.082- 0.484- O0.31- 0.14- 0.26-
>=-10-days-lost- O0.18- 0.017- O0.32- O0.03- 0.012-
Model-are-controlled-for-absence-from-school-
89
Figure 4.1. Growth curve model illustrating change in mean GPA over time. Note: Academic periods are centered around the quarter of injury. Two quarters prior to injury, the quarter when injury occurred, and two quarters following injury are shown. Dashed and dotted lines show 95% confidence intervals. Inset shows graph with a truncated y-axis to emphasis effects.
90
CHAPTER 5: CONCLUSION AND RECOMMENDATIONS
91
5.1 Summary of Findings
The research presented in this dissertation aims to increase the body of
knowledge regarding the post-acute neurocognitive impact of concussion injuries
in adolescent athletes. First, a systematic review of current literature addressing
the post-acute cognitive effects of concussion in adolescents found that
adolescent athletes experience a range of neurocognitive symptom deficits six or
more days following their injury. Athletes in the reviewed studies demonstrated
significant declines in the domains of reactions speed, memory, and total
symptom score during post-acute follow up.
The reviewed studies support the current understanding of adolescent
athlete recovery from concussion injury. While adults generally resolve
symptoms within the first week after injury, recent research indicates that young
athletes may begin showing new cognitive symptoms 7-10 days following
injury.(33, 35) The adolescent brain may be more sensitive to metabolic changes
following injury than adult brains, impacting recovery time and partially
explaining the delayed recovery trajectory observed in adolescents compared to
adults.(33, 50)
Descriptive analyses confirm that the study data are consistent with other
study populations and provide information about subpopulations, injury severity,
and comparative results for athletes with concussion injury and athletes with
ankle/leg injury. During this study period, 494 concussion injuries and 473 ankle
or leg injuries were recorded in the study school district. Many of the results
presented here are consistent with other studies examining concussion injuries
92
and recovery in high school athletes. First, football (36% of concussion injuries,
injury rate=5.19/10,000AE, 95%CI 4.42-5.96) and lacrosse (15% of concussion
injuries, injury rate= 4.77/10,000AE, 95%CI 3.68-5.89) account for a large
proportion of concussion injuries in this population. These result align with
results from a 10 year prospective study (between academic years 1997-1998 and
2007-2008) of concussion rates in the same study population that found high
rates of concussion in athletes participating in these collision sports (39).This
trend is true in other study populations, as well. (9, 45)
Additionally, concussion severity results in this study are consistent with
data from other populations.(45, 108, 109) In this study, football,
baseball/softball, basketball and wrestling had a higher proportion of concussion
injuries that resulted in more than 14 days of time lost from sport than
concussion injuries that resulted in less than 10 days of time lost.
Finally, in the subpopulation where gender could be inferred from the
sport played, female athletes account for a greater proportion of concussion in
each sport except lacrosse. Female athletes were also more likely to have more
severe concussion than male athletes participating in equivalent sports. These
trends are not evident when comparing athletes with ankle and leg injuries.
In this study population, there was a significantly higher proportion of
concussion injuries in younger athletes compared to ankle or leg injuries.
Athletes with concussion injuries were twice as likely to be in grade 9 or grade 10
than in grade 12 when compared to athletes with ankle and leg injuries. This
result is consistent with younger athletes being more susceptible to concussion
93
injuries (33, 36), but could also be explained by younger athletes being more
willing to openly report injury symptoms than older athletes.
Athletes in this study population with concussion injury had greater odds
of a significant increase in the number of days absent from school than athletes
with ankle or leg injuries. This result is of particular interest in the context of this
dissertation as absence from school is closely tied to academic performance.
Students who are absent from the classroom miss instructional time and are
more likely to perform poorly on classroom and standardized assessment.(110)
In longitudinal analyses, high school students with sport-related
concussion did not suffer a significant decline in academic performance
compared to athletes with ankle or leg injury. Athletes in both injury groups
experienced a decline in mean GPA during the quarter of injury compared to the
quarter prior to injury. When analyses were restricted to younger athletes
(grades 9 or 10) or those with more severe injuries (time lost from sport greater
than 10 days), athletes with concussion injury demonstrated a greater decline in
average grades over the observation period than athletes with ankle or leg injury.
There appears to be a significant decline in academic performance in
younger athletes (grades 9 or 10) with severe concussion injuries compared to
athletes with similarly severe ankle or leg injury (measured by time lost from
sport).
All adolescent athletes should be closely monitored following concussion
injury, but particular attention should be paid to the neurocognitive
rehabilitation and academic return of younger athletes. Concern over academic
94
time lost must be balanced with cognitive rest necessary for recovery from
concussion injures. Accommodation in the classroom may be necessary to
prevent negative effects on academic performance measures in athletes following
concussion injury.
5.2 Public Health Implications
An active lifestyle is associated with risk of injury. Physically active
individuals are more likely to suffer injuries as a result of their action.(121)
Sports and recreational pursuits are an important part of social and cultural
activity and also a source of risk. Participation in structured activity during
adolescence is associated with positive psychological, social, and educational
outcomes.(19-21) Evidence is emerging indicating that concussions are more
common and carry a heavier burden of consequences than previously
suspected.(11, 81)
There is potential for concussions to have significant impact on academic
performance in high school athletes. In adolescent athletes with concussion,
attention must be paid to balancing the required cognitive rest with academic
demands. Minor attention deficit or impaired information processing can have a
significant impact on an adolescent’s ability to function academically(58) and
evidence suggests that early return to school (before resolution of symptoms)
may prolong recovery.(33, 75)
Athletes with injuries miss more school than uninjured students, and
athletes with concussion are absent from school more than students with non-
95
concussion injuries. These absences result in a loss of instructional time that
impacts the student’s ability to perform academically. (110) This concern for loss
of instructional time is in conflict with the current policy of cognitive rest
following concussion injury. Return to the classroom must be balanced with the
need for cognitive rest when considering individual recovery plans.
Given the far reaching potential impact of adolescent concussion injuries
and the increasing public health burden, these injuries warrant the significant
attention and research they are receiving. Physical activity and participation in
athletics are important for adolescent development and should be encouraged.
Minimizing the negative effects of injuries that results from these activities will
ensure that adolescents maximize the potential benefits of sports participation.
5.3 Policy Implications
In November 2012, the 4th International Consensus Conference on
Concussion in Sport held in Zurich resulted in the publication of an updated
consensus statement building on previous statements and presenting an
improved understanding of concussion diagnosis and recovery management. The
group encourages investigations of the pathophysiologic mechanisms, symptom
identification and management, genetic markers, and neuropsychological
assessment of concussion injuries.(4) The statement also encourages extending
the recovery time for adolescent athletes due to differing physiological responses
and increased risk of diffuse cerebral swelling in children.
96
For any athletes with concussion, cognitive rest may be as important as
physical rest for recovery.(80, 81) Poor communication between healthcare
providers and school administration make it difficult for adolescent athletes to
get the rest they need to heal.(9, 33, 71, 81) Awareness of the impact of
concussions on academic outcomes may help with future identification of high-
risk athletes and allow educators opportunity for targeted intervention to prevent
negative academic consequences. Currently, many athletes are expected to
return to school environments that require focused attention over long periods
and exposure to loud environments within days of injury.(57, 82, 83)
Administrative policies allowing for cognitive recovery periods and individualized
academic rehabilitation plans will allow full recovery from injury, minimizing
academic consequences. The research presented here was conducted within a
school district that allocates significant resources to medical resources in their
athletics programs. This may have resulted in a shift on effects tested toward the
null, as district personnel may have been better educated about appropriate
accommodations for athletes following concussion injury.
As further research emerges outreach efforts should be made to educate
athletes, referees, parents, coaches, school administrators, and health care
providers on the prevention, detection, management, and potential consequences
of concussion injury. Young athletes and those with post-acute symptom
presentation should receive recovery care that is coordinated with academic
programs that includes cognitive therapy and academic support services.
97
Finding a balance between allowing for appropriate recovery and
rehabilitation that includes adequate attention to cognitive rest, maintenance of
social connections, and academic support is likely to be a challenging and highly
individual endeavor, but may result in better long term outcomes for high school
athletes with sport-related concussion injuries.
5.4 Limitations of Current Research
The research presented in this dissertation is an exploratory examination
of academic performance in high school athletes with concussion. This research
was limited by several factors that could be overcome in future research projects.
First, participation data for the study period was not available. Therefore, injury
rates presented in Chapter 3 are based on average participation numbers for the
two academic years following the study period. As participation numbers are
fairly stable over time in the study district, it is anticipated that the numbers
presented are an adequate estimate of injury rates. These rates align with other
study data.
Second, data on variables that may have been identifiers were restricted by
the school district. Access to calendar age, gender, history of concussion, and
other injury or medical history were not available for these analyses. Completion
of these data would make the results of analyses more meaningful. Some medical
conditions and learning disabilities impact concussion assessment and recovery.
98
In particular, history of concussion is a predictor of concussion recovery and
would have added to the value of the analyses presented.
The current research should be duplicated in other populations. The study
population was drawn from a large suburban school district that is has access to
resources that many smaller or less well funded schools may not have access to.
The study school district places substantial resources into athletics and athletic
injury management, accompanied by professional development for district
administration and staff about injury management and recovery requirements.
This awareness and level of support is not available in most school districts in the
country. Attention in this district to injury management and recovery may have
resulted in diminishing the effects being studied for this dissertation. Students in
this district may receive academic accommodations following injury that would
not be available to injured students in other school districts.
Additional studies would benefit from a prospective design with a common
assessment of academic performance to reduce the amount of error due to the
variability of outcome measurements noted in this study.
5.5 Future Research Considerations.
The research presented in this dissertation could be expanded on by
duplication in another study population, examination of additional variables
measuring academic outcomes such as standardized academic assessments, or by
implementing a prospective longitudinal study.
99
Concussion research, particularly in adolescent populations, is of great
current interest. Policy and administrative support for research should be
leveraged to launch additional studies while interest is high. While biological,
imaging, and genetic studies are of value in concussion research, population
based research studies have the greatest potential for actually informing how
concussion injuries are managed and minimizing the long term impact of these
injuries.
100
LIST OF REFERENCES
101
List of References
1. Aubry M, Cantu R, Dvorak J, Graf-Baumann T, Johnston K, Kelly J, et al. Summary and agreement statement of the first International Conference on Concussion in Sport, Vienna 2001. British Journal of Sports Medicine. 2002;36(1):6-7
2. McCrory P, Johnston K, Meeuwisse W, Aubry M, Cantu R, Dvorak J, et al. Summary and agreement statement of the 2nd International Conference on Concussion in Sport, Prague 2004. British Journal of Sports Medicine. 2005;39(suppl 1):i78-i86
3. McCrory P, Meeuwisse W, Johnston K, Dvorak J, Aubry M, Molloy M, et al. Consensus Statement on Concussion in Sport–the 3rd International Conference on Concussion in Sport held in Zurich, November 2008. South African Journal of Sports Medicine. 2009;21(2).
4. McCrory P, Meeuwisse WH, Aubry M, Cantu RC, Dvorak J, Echemendia RJ, et al. Consensus Statement on Concussion in Sport: The 4th International Conference on Concussion in Sport, Zurich, November 2012. Journal of Athletic Training. 2013;48(4):554-75.
5. Daneshvar DH, Nowinski CJ, McKee AC, Cantu RC. The Epidemiology of Sport-Related Concussion. Clinics in sports medicine. 2011;30(1):1-17.
6. Broglio SP, Guskiewicz KM. Concussion in Sports: The Sideline Assessment. Sports Health: A Multidisciplinary Approach. 2009;1(5):361-9
7. Collie A, Makdissi M, Maruff P, Bennell K, McCrory P. Cognition in the days following concussion: comparison of symptomatic versus asymptomatic athletes. Journal of Neurology, Neurosurgery & Psychiatry. 2006;77(2):241-5.
8. Hanna-Pladdy B, Berry ZM, Bennett T, Phillips HL, Gouvier WD. Stress as a Diagnostic Challenge for Postconcussive Symptoms: Sequelae of Mild Traumatic Brain Injury or Physiological Stress Response. The Clinical Neuropsychologist. 2001;15(3):289.
9. Meehan WP, Bachur RG. Sport-Related Concussion. Pediatrics. 2009;123(1):114-23.
102
10. Randolph C, Millis S, Barr WB, McCrea M, Guskiewicz KM, Hammeke TA, et al. Concussion Symptom Inventory: an empirically derived scale for monitoring resolution of symptoms following sport-related concussion. Archives of Clinical Neuropsychology. 2009;24(3):219.
11. Bailes JE, Cantu RC. Head injury in athletes. Neurosurgery. 2001;48(1):26.
12. Grindel SH, Lovell MR, Collins MW. The Assessment of Sport-Related Concussion: The Evidence Behind Neuropsychological Testing and Management. Clinical Journal of Sport Medicine. 2001;11(3):134-43.
13. Hinton-Bayre AD, Geffen GM, Geffen LB, McFarland KA, Frijs P. Concussion in Contact Sports: Reliable Change Indices of Impairment and Recovery. Journal of Clinical and Experimental Neuropsychology. 1999;21(1):70
14. Maddocks DL, Dicker GD, Saling MM. The assessment of orientation following concussion in athletes. Clinical Journal of Sport Medicine. 1995;5(1):32.
15. McCrea M, Kelly J, Randolph C. The Standardized Assessment of Concussion: Manual for Administration, Scoring and Interpretation: Brain Injury Assoc.; 1997.
16. McCrea M, Kelly JP, Randolph C, Kluge J, Bartolic E, Finn G, et al. Standardized assessment of concussion (SAC): on-site mental status evaluation of the athlete. The Journal of Head Trauma Rehabilitation. 1998;13(2):27.
17. Jiang J-Y, Gao G-Y, Li W-P, Yu M-K, Zhu C. Early Indicators of Prognosis in 846 Cases of Severe Traumatic Brain Injury. Journal of Neurotrauma. 2002;19(7):869-74.
18. Johnston KM, McCrory P, Mohtadi NG, Meeuwisse W. Evidence-Based Review of Sport-Related Concussion: Clinical Science. Clinical Journal of Sport Medicine. 2001;11(3):150-9.
19. Barber B, Eccles J, Stone M. Whatever Happened to the Jock, the Brain, and the Princess? Journal of Adolescent Research. 2001;16(5):429.
103
20. Eccles J, Barber B, Stone M, Hunt J. Extracurricular activities and adolescent development. Journal of Social Issues. 2003;59(4):865-9.
21. Sallis JF. Leisure-time Physical Activity and Depression in Adolescence. Clinical Journal of Sport Medicine. 2011;21(1):72.
22. Romer D. Adolescent risk taking, impulsivity, and brain development: Implications for prevention. Developmental Psychobiology. 2010;52(3):263-76.
23. Spear L. The psychobiology of adolescence. Authoritative Communities. 2008:263-80.
24. Bailey R. Evaluating the relationship between physical education, sport and social inclusion. Educational Review. 2005;57(1):71–90.
25. Bailey R. Physical education and sport in schools: A review of benefits and outcomes. Journal of School Health. 2006;76(8):397–401.
26. Darling N, Caldwell L, Smith R. Participation in School-Based Extracurricular Activities and Adolescent Adjustment. Journal of Leisure Research. 2005;37(1):51-77.
27. Lerner R, Lerner J, De Stefanis I, Apfel A. Understanding Developmental Systems in Adolescence. Journal of Adolescent Research. 2001;16(1):9.
28. Trudeau F, Shephard RJ. Physical education, school physical activity, school sports and academic performance. International Journal of Behavioral Nutrition and Physical Activity. 2008;5(1):10.
29. Eccles JS, Barber BL, Stone M, Hunt J. Extracurricular Activities and Adolescent Development. Journal of Social Issues. 2003;59(4):865-89.
30. Bergeron MF. The Young Athlete: Challenges of Growth, Development, and Society. Current Sports Medicine Reports. 2010;9(6):356.
31. Farrey T. Competitive youth sports in society: what President Obama needs to know to get-and keep-kids moving. Current Sports Medicine Reports. 2010;9(6):359.
104
32. Manuel JC, Shilt JS, Curl WW, Smith JA, DuRant RH, Lester L, et al. Coping with sports injuries: An examination of the adolescent athlete. Journal of Adolescent Health. 2002;31:391-3.
33. Grady MF. Concussion in the Adolescent Athlete. Current Problems in Pediatric and Adolescent Health Care. 2010;40(7):154-69.
34. Rogol A, Roemmich J, Clark P. Growth at puberty. The Journal of adolescent health: official publication of the Society for Adolescent Medicine. 2002;31(6 Suppl):192.
35. Lovell MR, Collins MW, Iverson GL, Field M, Maroon JC, Cantu R, et al. Recovery from mild concussion in high school athletes. Journal of Neurosurgery. 2003;98(2):296–301.
36. Comstock R, Knox C, Yard E, Gilchrist J. Sports-related injuries among high school athletes-United States, 2005-06 school year. Morbidity and Mortality Weekly Report. 2006;55(38):1037-40.
37. Gessel LM, Fields SK, Collins CL, Dick RW, Comstock RD. Concussions Among United States High School and Collegiate Athletes. Journal of Athletic Training. 2007;42(4):495-503.
38. Gilchrist J, Thomas K, Wald M, Langlois J. Nonfatal traumatic brain injuries from sports and recreation activities-United States, 2001-2005. Morbidity and Mortality Weekly Report. 2007;56(29):733-7.
39. Lincoln AE, Caswell SV, Almquist JL, Dunn RE, Norris JB, Hinton RY. Trends in Concussion Incidence in High School Sports: A Prospective 11-Year Study. The American Journal of Sports Medicine. 2011;published online before print January 29, 2011.
40. Swenson DM, Yard EE, Fields SK, Comstock RD. Patterns of recurrent injuries among US high school athletes, 2005ñ2008. The American Journal of Sports Medicine. 2009;37(8):1586.
41. 2010-2011 high school athletics participation survey.2012 April 2012.
105
42. Prevention CfDC. Sports-related recurrent brain injuries--United States. MMWR: Morbidity and Mortality Weekly Report. 1997;46(10):224-7.
43. Schatz P, Moser RS. Current Issues in Pediatric Sports Concussion. The Clinical Neuropsychologist. 2011;25(6).
44. Lincoln AE, Caswell SV, Almquist JL, Dunn RE, Norris JB, Hinton RY. Trends in concussion incidence in high school sports a prospective 11-year study. The American Journal of Sports Medicine. 2011;39(5):958-63.
45. Marar M, McIlvain NM, Fields SK, Comstock RD. Epidemiology of concussions among United States high school athletes in 20 sports. The American Journal of Sports Medicine. 2012;40(4):747-55.
46. Hootman JM, Dick R, Agel J. Epidemiology of Collegiate Injuries for 15 Sports: Summary and Recommendations for Injury Prevention Initiatives. Journal of Athletic Training. 2007;42(2):311-9.
47. Thurman DJ, Alverson C, Dunn KA, Guerrero J, Sniezek JE. Traumatic Brain Injury in the United States: A Public Health Perspective. Journal of Head Trauma Rehabilitation. 1999;14(6):602-15.
48. Rutherford W, Merrett J, Mcdonali J. Sequelae of concussion caused by minor head injuries. The Lancet. 1977;309(8001):1-4.
49. Symonds SC. Concussion and its Sequelae. The Lancet. 1962;279(7219):1-5.
50. Pickering A, Townend W. 3: A Study of Outcomes Following Head Injury Among Children and Young Adults in Full-Time Education. Annals of Emergency Medicine. 2008;51(4):471-.
51. Collins MW, Grindel SH, Lovell MR, Dede DE, Moser DJ, Phalin BR, et al. Relationship between concussion and neuropsychological performance in college football players. Jama. 1999;282(10):964.
52. Ellemberg D, Henry LC, Macciocchi SN, Guskiewicz KM, Broglio SP. Advances in Sport Concussion Assessment: From Behavioral to Brain Imaging Measures. Journal of Neurotrauma. 2009;26(12):2365-82.
106
53. Langlois JA, Rutland-Brown W, Wald M. The Epidemiology and Impact of Traumatic Brain Injury: A Brief Overview. Journal of Head Trauma Rehabilitation. 2006;21(5):375-8
54. Covassin T, Elbin RJ, Harris W, Parker T, Kontos AP. The Role of Age and Sex in Symptoms, Neurocognitive Performance, and Postural Stability in Athletes After Concussion. American Journal of Sports Medicine. 2012;Published online first. Epub April 26, 2012.
55. Granite V, Carroll J. Psychological response to atheltic injury: sex differences. Journal of Sport Behavior. 2002;25:546-53.
56. Weiss MR. Psychological aspects of sport-injury rehabilitation: A developmental perspective. Journal of Athletic Training. 2003;38(2):172.
57. Casey B, Jones RM, Hare TA. The adolescent brain. Annals of the New York Academy of Sciences. 2008;1124(1):111-26.
58. McCrory P, Collie A, Anderson V, Davis G. Can we manage sport related concussion in children the same as in adults? British Journal of Sports Medicine. 2004;38(5):516.
59. Anderson V, Catroppa C, Morse S, Haritou F, Rosenfeld J. Outcome from mild head injury in young children: a prospective study. Journal of Clinical and Experimental Neuropsychology. 2001;23(6):705-17.
60. Ommaya A, Goldsmith W, Thibault L. Biomechanics and neuropathology of adult and paediatric head injury. British Journal of Neurosurgery. 2002;16(3):220-42.
61. Ponsford J, Willmott C, Rothwell A, Cameron P, Ayton G, Nelms R, et al. Cognitive and behavioral outcome following mild traumatic head injury in children. The Journal of Head Trauma Rehabilitation. 1999;14(4):360.
62. Bleiberg J, Garmoe W, Halpern E, Reeves D, Nadler J. Consistency of within-day and across-day performance after mild brain injury. Cognitive and Behavioral Neurology. 1997;10(4):247.
107
63. Makdissi M, Collie A, Maruff P, Darby D, Bush A, McCrory P, et al. Computerised cognitive assessment of concussed Australian Rules footballers. British Journal of Sports Medicine. 2001;35(5):354.
64. Iverson GL, Gaetz M, Lovell MR, Collins MW. Cumulative effects of concussion in amateur athletes. Brain Injury. 2004;18(5):433-43.
65. Gronwall D. Cumulative and persisting effects of concussion on attention and cognition. Mild head injury. 1989:153-62.
66. Gronwall D, Wrightson P. Cumulative effect of concussion. The Lancet. 1975;306(7943):995-7.
67. Matser E, Kessels A, Lezak M, Jordan B, Troost J. Cumulative problems with memory and planning in amateur soccer. Journal of the American Medical Association. 1999;282:971-3.
68. Moser RS, Schatz P, Jordan BD. Prolonged Effects of Concussion in High School Athletes. Neurosurgery. 2005;57(2):300-6.
69. Bloom GA, Loughead TM, Shapcott EJB, Johnston KM, Delaney JS. The prevalence and recovery of concussed male and female collegiate athletes. European Journal of Sport Science. 2008;8(5):295-303.
70. Bailey CM, Samples HL, Broshek DK, Freeman JR, Barth JT. The Relationship Between Psychological Distress and Baseline Sports-Related Concussion Testing. Clinical Journal of Sport Medicine. 2010;20(4):272-7.
71. Bleiberg J, Cernich AN, Cameron K, Sun W, Peck K, Ecklund LTCJ, et al. Duration of Cognitive Impairment After Sports Concussion. Neurosurgery. 2004:1073-80
72. Dick RW. Is there a gender difference in concussion incidence and outcomes? British Journal of Sports Medicine. 2009;43(Suppl_1):i46-i50.
73. Meehan III WP, Bachur RG. Sport-related concussion. Pediatrics. 2009;123(1):114.
108
74. McCrea M, Iverson GL, Echemendia RJ, Makdissi M, Raftery M. Day of injury assessment of sport-related concussion. British Journal of Sports Medicine. 2013;47(5):1-14.
75. Sarmiento K, Mitchko J, Klein C, Wong S. Evaluation of the Centers for Disease Control and Prevention's Concussion Initiative for High School Coaches:ìHeads Up: Concussion in High School Sportsî. Journal of School Health. 2010;80(3):112-8.
76. Berger-Gross P, Shackelford M. Closed-head injury in children: neuropsychological and scholastic outcomes. Perceptual and motor skills. 1985;61(1):254.
77. Ewing-Cobbs L, Fletcher JM, Levin HS, Iovino I, Miner ME. Academic Achievement and Academic Placement Following Traumatic Brain Injury in Children and Adolescents: A Two-Year Longitudinal Study. Journal of Clinical and Experimental Neuropsychology. 1998;20(6):769-81.
78. Taylor HG, Swartwout MD, Yeates KO, Walz NC, Stancin T, Wade SL. Traumatic brain injury in young children: postacute effects on cognitive and school readiness skills. Journal of the International Neuropsychological Society. 2008;14(05):734–45.
79. Gerberich SG, Gibson RW, Fife D, Mandel JS, Aeppli D, Le CT, et al. Effects of brain injury on college academic performance. Neuroepidemiology. 1997;16(1):1-14.
80. Majerske CW, Mihalik JP, Ren D, Collins MW, Reddy CC, Lovell MR, et al. Concussion in Sports: Postconcussive Activity Levels, Symptoms, and Neurocognitive Performance. Journal of Athletic Training. 2008;43(3):265-74.
81. Powell JW, Barber-Foss KD. Traumatic Brain Injury in High School Athletes. JAMA. 1999;282(10):958-63.
82. Kutcher JS, Eckner JT. At-Risk Populations in Sports-Related Concussion. Current Sports Medicine Reports. 2010;9(1):16-20.
83. Reddy CC, Collins MW. Sports Concussion: Management and Predictors of Outcome. Current Sports Medicine Reports. 2009;8(1):10-5.
109
84. Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, et al. Meta-analysis of observational studies in epidemiology. JAMA: The Journal of the American Medical Association. 2000;283(15):2008-12.
85. Petitti DB. Meta-analysis, decision analysis, and cost-effectiveness analysis: methods for quantitative synthesis in medicine: Oxford University Press US; 2000.
86. Sutton AJ. Methods for meta-analysis in medical research. Michigan, US: John Wiley; 2000.
87. Petitti DB. Approaches to heterogeneity in meta-analysis. Statistics in Medicine. 2001;20(23):3625-33.
88. Rothstein HR, Sutton AJ, Borenstein M. Publication bias in meta-analysis: Prevention, assessment and adjustments: Wiley. com; 2006.
89. Collins MW, Field M, Lovell MR, Iverson GL, Johnston KM, Maroon J, et al. Relationship Between Postconcussion Headache and Neuropsychoogical Test Performance in High School Athletes. American Journal of Sports Medicine. 2003;31(2):168-73.
90. Moser RS, Glatts C, Schatz P. Efficacy of Immediate and Delayed Cognitive and Physical Rest for Treatment of Sports-Related Concussion. Journal of Pediatrics. 2012;In Press. Epub May 2012.
91. Lovell MR, Collins MW, Iverson GL, Johnston KM, Bradley JP. Grade 1 or ìdingî concussions in high school athletes. The American Journal of Sports Medicine. 2004;32(1):47.
92. Iverson GL, Lovell M, Collins MW. Immediate Post-Concussion Assessment and Cognitive Testing (ImPACT) Normative Data. Pittsburgh, Pa: University of Pittsburgh; 2003.
93. McClincy MP, Lovell MR, Pardini J, Collins MW, Spore MK. Recovery from sports concussion in high school and collegiate athletes. Brain Injury. 2006;20(1):33-9.
110
94. Maugans TA, Farley C, Altaye M, Leach J, Cecil KM. Pediatric Sports-Related Concussion Produces Cerebral Blood Flow Alterations. Pediatrics. 2012;129(1):1-10.
95. Thomas DG, Collins MW, Saladino RA, Frank V, Raab J, Zuckerbraun NS. Identifying Neurocognitive Deficits in Adolescents Following Concussion. Academy of Emergency Medicine. 2011;18(3):246-54.
96. Teasdale TW, A E. Duration of cognitive dysfunction after concussion, and cognitive dysfunction as a risk factor: a population study of young men. BMJ. 1997;315:569.
97. Field M, Collins MW, Lovell MR, Maroon J. Does age play a role in recovery from sports-related concussion? A comparison of high school and collegiate athletes. Journal of Pediatrics. 2003;142:546–53.
98. Sim A, Terryberry-Spohr L, Wilson KR. Prolonged recovery of memory functioning after mild traumatic brain injury in adolescent athletes. Journal of Neurosurgery. 2008;108(3):511-6.
99. Moser RS, Schatz P. Enduring effects of concussion in youth athletes. Archives of Clinical Neuropsychology. 2002;17(1):91-100.
100. Kontos AP, Braithwaite R, Dakan S, Elbin RJ. Computerized Neurocognitive Testing within 1 Week of Sport-Related Concussion: Meta-analytic Review and Analysis of Moderating Factors. Journal of the International Neuropsychological Society. 2014;20(03):324-32.
101. Dougan BK, Horswill MS, Geffen GM. Athletes’ Age, Sex, and Years of Education Moderate the Acute Neuropsychological Impact of Sports-Related Concussion: A Meta-analysis. Journal of the International Neuropsychological Society. 2013;FirstView:1-17.
102. Guskiewicz KM, Weaver NL, Padua DA, Garrett Jr WE. Epidemiology of concussion in collegiate and high school football players. The American Journal of Sports Medicine. 2000;28(5):643.
111
103. Kirkwood MW, Yeates KO, Taylor HG, Randolph C, McCrea M, Anderson VA. Management of pediatric mild traumatic brain injury: A neuropsychological review from injury through recovery. The Clinical Neuropsychologist. 2008;22(5):769.
104. Scott WW. Physiology of concussion. Archives of Neurology and Psychiatry. 1940;43(2):270.
105. Terrell TR. Concussion in athletes. Southern medical journal. 2004;97(9):837-42.
106. United States Census Bureau: American Community Survey: United States Census Bureau; [updated February 22, 2011February 11, 2011]. Available from: http://www.census.gov/acs/www/.
107. [BLINDED] County Public Schools - Statistics: [BLINDED] County Public Schools; [updated August 11, 2010February 11, 2011]. Available from: http://www.[BLINDED].edu/about.htm.
108. Chrisman SP, Rivara FP, Schiff MA, Zhou C, Comstock RD. Risk factors for concussive symptoms 1 week or longer in high school athletes. Brain Injury. 2013;27(1):1-9.
109. Meehan III WP, d'Hemecourt P, Comstock RD. High School Concussion in the 2008-2009 Academic Year: Mechanisms, Symptoms, and Management. Am J Sports Med. 2010;38(12):2405-9.
110. Moonie S, Sterling DA, Figgs LW, Castro M. The Relationship Between School Absence, Academic Performance, and Asthma Status. Journal of School Health. 2008;78(3):140-8.
111. McCrea M, Hammeke T, Olsen G, Leo P, Guskiewica K. Unreported Concussion in High School Football Players: Implications for Prevention. Clinical Journal of Sports Medicine. 2004;14(1):13-7.
112. Schulz MR, Marshall SW, Mueller FO, Yang J, Weaver NL, Kalsbeek WD, et al. Incidence and Risk Factors for Concussion in High School Athletes, North Carolina, 1996–1999. American Journal of Epidemiology. 2004;160(10):937-44.
112
113. Parker JDA, Creque RE, Barnhart DL, Harris JI, Majeski SA, Wood LM, et al. Academic achievement in high school: does emotional intelligence matter? Personality and Individual Differences. 2004;37(7):1321-30.
114. Barnes P, Price L, Maddocks A, Lyons R, Nash P, M M. Unnecessary school absence after minor injury: case-control study. British Medical Journal. 2001;323(7320):1034-5.
115. Hyman J, Jewetz S, Matsumoto H, Choe J, Vitale M. Risk factors for school absence after acute orthopaedic injury in New York City. Journal of Pediatric Orthopedics. 2007;27(4):415-20.
116. Sesko A, Choe J, Vitale M, Ugwonali O, JE H. Pediatric orthopaedic injuries: the effect of treatment on school attendance. Journal of Pediatric Orthopedics. 2005;25(5):661-5.
117. Buzzini SRR, Guskiewicz KM. Sport-related concussion in the young athlete. Current Opinion in Pediatrics. 2006;18(4):376-82.
118. Stringer GA. Effects of Sport-Related Concussion on Academic Performance in High School Athletes. [Unpublished Doctoral Dissertation]. In press 2014.
119. Nakagawa S, Cuthill IC. Effect size, confidence interval and statistical significance: a practical guide for biologists. Biological Reviews. 2007;82(4):591-605.
120. Lee J. Tripartite Growth Trajectories of Reading and Math Achievement Tracking National Academic Progress at Primary, Middle, and High School Levels. American Educational Research Journal. 2010;47(4):800-32.
121. Robertson LS. Injury Epidemiology: Research and Control Strategies. New York, New York: Oxford University Press; 2007.
113
VITA &Education&PhD- Drexel-University-School-of-Public-Health,-Philadelphia,-PA- Epidemiology- -
MPH- Drexel-University-School-of-Public-Health,-Philadelphia,-PA- Epidemiology-
BA- Bryn-Mawr-College,-Bryn-Mawr,-PA- - - - Chemistry-
Academic&and&Research&Appointments&Teaching-Assistant- - - - - - - 2010O2011-
Drexel!University!School!of!Public!Health!! !! ! ! Philadelphia,!PA!Research-Associate-- - - - - - - 2009O2011-
Thomas!Jefferson!University,!Division!of!Biostatistics! !! ! Philadelphia,!PA!Honors&and&Awards&Drexel-University-Scholarship- - - - - - 2009O2011-
Hygeia-Academic-Honor-Society-- - - - - 2009-
Bryn-Mawr-College-Merit-Scholarship- - - - - 1995O2000-
Publications&Journal-articles-
1.-Peck-AR,-Witkiewicz-AK,-Chengbao-L,-Stringer-GA,-Klimowicz-AC,-Pequignot-E,-Freydin-B,-Tran-
TH,-Yang-N,-Rosenberg-AL,-Hooke-JA,-Kovatich-AJ,-Nevalainen-MT,-Shriver-CD,-Hyslop-T,-Sauter-G,-
Rimm-DL,-Magliocco-AM,-Rui-H.-Loss-of-Nuclear-Localized-and-Tyrosine-Phosphorylated-Stat5-in-
Breast-Cancer-Predicts-Poor-Clinical-Outcomes-and-Increased-Risk-of-Antiestrogen-Therapy-
Failure.-(Journal-of-Clinical-Oncology,-accepted-for-publication)-
2.-Stringer-GA,-Mieja-A,-Rui-H,-Mitchell-E,-Michael-YL,-Hyslop-T.-Serum-25OHydroxyvitamin-D-and-
Breast-Cancer-Risk:-a-metaOanalysis.-(in-review-by-Journal-of-Clinical-Oncology)-
PeerOreviewed-Abstracts--
1.-Peck-AR,-Witkiewicz-AK,-Chengbao-L,-Stringer-GA,-Klimowicz-AC,-Pequignot-E,-Freydin-B,-Tran-
TH,-Yang-N,-Rosenberg-AL,-Hooke-JA,-Kovatich-AJ,-Nevalainen-MT,-Shriver-CD,-Hyslop-T,-Sauter-G,-
Rimm-DL,-Magliocco-AM,-Rui-H.-Loss-of-Nuclear-Localized-and-Tyrosine-Phosphorylated-Stat5:-A-
Predictor-of-Poor-Clinical-Outcome-and-Increased-Risk-of-Antiestrogen-Therapy-Failure-in-Breast-
Cancer.-Abstract-(accepted),-American-Association-of-Cancer-Research-Annual-Conference,-2010-
in-Orlando,-Florida-
2.-Stringer-GA,-Lee-BK.-Recent-Injury-and-Alcohol-Use-Behavior-in-High-School-Athletes.-Abstract-
(accepted),-Society-for-Epidemiologic-Research,-Annual-Meeting,-June-2012,-Minneapolis,-MN-
Teaching&Department-of-Epidemiology-and-Biostatistics,-Drexel-University-School-of-Public-Health-
Instructor:-Epidemiology-in-Public-Health-(3-credits,-undergraduate),-Content-development-and-
instruction-(Spring-2011--
Teaching-Assistant:-Introduction-to-Descriptive-Epidemiology-and-Biostatistics-(3-credits,-
graduate-online),-Instruction-and-grading-(Fall-2011,-Winter-2010-
Teaching-Assistant:-Introduction-to-Analytical-Epidemiology-and-Biostatistics-(3-credits,-graduate-
online),-Instruction-and-grading-(Winter-2011,-Spring-2010-
Teaching-Assistant:-Design-and-Analysis-of-Epidemiological-Studies-(3-credits,-graduate-online),-
Instruction-and-grading-(Spring-2012)-
Teaching-Assistant:-Intermediate-Epidemiology-(3-credits,-graduate),-Instruction-and-grading-
(Fall-2010)-
-